Sepsis is a life-threatening condition characterized by an exaggerated immune response to pathogens, leading to organ damage and high mortality rates in the intensive care unit. Although deep learning has achieved impressive performance on prediction and classification tasks in medicine, it requires large amounts of data and lacks explainability, which hinder its application to sepsis diagnosis. We introduce a deep learning framework, called scCaT, which blends the capsulating architecture with Transformer to develop a sepsis diagnostic model using single-cell RNA sequencing data and transfers it to bulk RNA data. The capsulating architecture effectively groups genes into capsules based on biological functions, which provides explainability in encoding gene expressions. The Transformer serves as a decoder to classify sepsis patients and controls. Our model achieves high accuracy with an AUROC of 0.93 on the single-cell test set and an average AUROC of 0.98 on seven bulk RNA cohorts. Additionally, the capsules can recognize different cell types and distinguish sepsis from control samples based on their biological pathways. This study presents a novel approach for learning gene modules and transferring the model to other data types, offering potential benefits in diagnosing rare diseases with limited subjects.

MOTIVATION: The confusion of acute inflammation infected by virus and bacteria or noninfectious inflammation will lead to missing the best therapy occasion resulting in poor prognoses. The diagnostic model based on host gene expression has been widely used to diagnose acute infections, but the clinical usage was hindered by the capability across different samples and cohorts due to the small sample size for signature training and discovery. RESULTS: Here, we construct a large-scale dataset integrating multiple host transcriptomic data and analyze it using a sophisticated strategy which removes batch effect and extracts the common information from different cohorts based on the relative expression alteration of gene pairs. We assemble 2680 samples across 16 cohorts and separately build gene pair signature (GPS) for bacterial, viral, and noninfected patients. The three GPSs are further assembled into an antibiotic decision model (bacterial-viral-noninfected GPS, bvnGPS) using multiclass neural networks, which is able to determine whether a patient is bacterial infected, viral infected, or noninfected. bvnGPS can distinguish bacterial infection with area under the receiver operating characteristic curve (AUC) of 0.953 (95% confidence interval, 0.948-0.958) and viral infection with AUC of 0.956 (0.951-0.961) in the test set (N = 760). In the validation set (N = 147), bvnGPS also shows strong performance by attaining an AUC of 0.988 (0.978-0.998) on bacterial-versus-other and an AUC of 0.994 (0.984-1.000) on viral-versus-other. bvnGPS has the potential to be used in clinical practice and the proposed procedure provides insight into data integration, feature selection and multiclass classification for host transcriptomics data. AVAILABILITY AND IMPLEMENTATION: The codes implementing bvnGPS are available at https://github.com/Ritchiegit/bvnGPS. The construction of iPAGE algorithm and the training of neural network was conducted on Python 3.7 with Scikit-learn 0.24.1 and PyTorch 1.7. The visualization of the results was implemented on R 4.2, Python 3.7, and Matplotlib 3.3.4.

MOTIVATION: Human gut microbiota plays a vital role in maintaining body health. The dysbiosis of gut microbiota is associated with a variety of diseases. It is critical to uncover the associations between gut microbiota and disease states as well as other intrinsic or environmental factors. However, inferring alterations of individual microbial taxa based on relative abundance data likely leads to false associations and conflicting discoveries in different studies. Moreover, the effects of underlying factors and microbe-microbe interactions could lead to the alteration of larger sets of taxa. It might be more robust to investigate gut microbiota using groups of related taxa instead of the composition of individual taxa. RESULTS: We proposed a novel method to identify underlying microbial modules, i.e. groups of taxa with similar abundance patterns affected by a common latent factor, from longitudinal gut microbiota and applied it to inflammatory bowel disease (IBD). The identified modules demonstrated closer intragroup relationships, indicating potential microbe-microbe interactions and influences of underlying factors. Associations between the modules and several clinical factors were investigated, especially disease states. The IBD-associated modules performed better in stratifying the subjects compared with the relative abundance of individual taxa. The modules were further validated in external cohorts, demonstrating the efficacy of the proposed method in identifying general and robust microbial modules. The study reveals the benefit of considering the ecological effects in gut microbiota analysis and the great promise of linking clinical factors with underlying microbial modules. AVAILABILITY AND IMPLEMENTATION: https://github.com/rwang-z/microbial_module.git.

The advances in single-cell RNA sequencing (scRNA-seq) technologies enable the characterization of transcriptomic profiles at the cellular level and demonstrate great promise in bulk sample analysis thereby offering opportunities to transfer gene signature from scRNA-seq to bulk data. However, the gene expression signatures identified from single cells are typically inapplicable to bulk RNA-seq data due to the profiling differences of distinct sequencing technologies. Here, we propose single-cell pair-wise gene expression (scPAGE), a novel method to develop single-cell gene pair signatures (scGPSs) that were beneficial to bulk RNA-seq classification to transfer knowledge across platforms. PAGE was adopted to tackle the challenge of profiling differences. We applied the method to acute myeloid leukemia (AML) and identified the scGPS from mouse scRNA-seq that allowed discriminating between AML and control cells. The scGPS was validated in bulk RNA-seq datasets and demonstrated better performance (average area under the curve [AUC] = 0.96) than the conventional gene expression strategies (average AUC$\le$ 0.88) suggesting its potential in disclosing the molecular mechanism of AML. The scGPS also outperformed its bulk counterpart, which highlighted the benefit of gene signature transfer. Furthermore, we confirmed the utility of scPAGE in sepsis as an example of other disease scenarios. scPAGE leveraged the advantages of single-cell profiles to enhance the analysis of bulk samples revealing great potential of transferring knowledge from single-cell to bulk transcriptome studies.

MOTIVATION: Hepatocellular carcinoma (HCC) is a primary malignancy with a poor prognosis. Recently, multi-omics molecular-level measurement enables HCC diagnosis and prognosis prediction, which is crucial for early intervention of personalized therapy to diminish mortality. Here, we introduce a novel strategy utilizing DNA methylation and RNA expression data to achieve a multi-omics gene pair signature (GPS) for HCC discrimination. RESULTS: The immune genes with negative correlations between expression and promoter methylation are enriched in the highly connected cancer-related pathway network, which are considered as the candidates for HCC detection. After that, we separately construct a methylation GPS (mGPS) and an expression GPS (eGPS), and then assemble them as a meGPS with five gene pairs, in which the significant methylation and expression changes occur between HCC tumor and non-tumor groups. Reliable performance has been validated by independent tissue (age, gender and etiology) and blood datasets. This study proposes a procedure for multi-omics GPS identification and develops a novel HCC signature using both methylome and transcriptome data, suggesting potential molecular targets for the detection and therapy of HCC. AVAILABILITY AND IMPLEMENTATION: Models are available at https://github.com/bioinformaticStudy/meGPS.git. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

High-throughput sequencing can detect tens of thousands of genes in parallel, providing opportunities for improving the diagnostic accuracy of multiple diseases including sepsis, which is an aggressive inflammatory response to infection that can cause organ failure and death. Early screening of sepsis is essential in clinic, but no effective diagnostic biomarkers are available yet. Here, we present a novel method, Recurrent Logistic Regression, to identify diagnostic biomarkers for sepsis from the blood transcriptome data. A panel including five immune-related genes, LRRN3, IL2RB, FCER1A, TLR5, and S100A12, are determined as diagnostic biomarkers (LIFTS) for sepsis. LIFTS discriminates patients with sepsis from normal controls in high accuracy (AUROC = 0.9959 on average; IC = [0.9722-1.0]) on nine validation cohorts across three independent platforms, which outperforms existing markers. Our analysis determined an accurate prediction model and reproducible transcriptome biomarkers that can lay a foundation for clinical diagnostic tests and biological mechanistic studies.

The superior performance of machine-learning scoring functions for docking has caused a series of debates on whether it is due to learning knowledge from training data that are similar in some sense to the test data. With a systematically revised methodology and a blind benchmark realistically mimicking the process of prospective prediction of binding affinity, we have evaluated three broadly used classical scoring functions and five machine-learning counterparts calibrated with both random forest and extreme gradient boosting using both solo and hybrid features, showing for the first time that machine-learning scoring functions trained exclusively on a proportion of as low as 8% complexes dissimilar to the test set already outperform classical scoring functions, a percentage that is far lower than what has been recently reported on all the three CASF benchmarks. The performance of machine-learning scoring functions is underestimated due to the absence of similar samples in some artificially created training sets that discard the full spectrum of complexes to be found in a prospective environment. Given the inevitability of any degree of similarity contained in a large dataset, the criteria for scoring function selection depend on which one can make the best use of all available materials. Software code and data are provided at https://github.com/cusdulab/MLSF for interested readers to rapidly rebuild the scoring functions and reproduce our results, even to make extended analyses on their own benchmarks.

Moonlighting proteins provide more options for cells to execute multiple functions without increasing the genome and transcriptome complexity. Although there have long been calls for computational methods for the prediction of moonlighting proteins, no method has been designed for determining moonlighting long noncoding ribonucleicacidz (RNAs) (mlncRNAs). Previously, we developed an algorithm MoonFinder for the identification of mlncRNAs at the genome level based on the functional annotation and interactome data of lncRNAs and proteins. Here, we update MoonFinder to MoonFinder v2.0 by providing an extensive framework for the detection of protein modules and the establishment of RNA-module associations in human. A novel measure, moonlighting coefficient, was also proposed to assess the confidence of an ncRNA acting in a moonlighting manner. Moreover, we explored the expression characteristics of mlncRNAs in sepsis, in which we found that mlncRNAs tend to be upregulated and differentially expressed. Interestingly, the mlncRNAs are mutually exclusive in terms of coexpression when compared to the other lncRNAs. Overall, MoonFinder v2.0 is dedicated to the prediction of human mlncRNAs and thus bears great promise to serve as a valuable R package for worldwide research communities (https://cran.r-project.org/web/packages/MoonFinder/index.html). Also, our analyses provide the first attempt to characterize mlncRNA expression and coexpression properties in adult sepsis patients, which will facilitate the understanding of the interaction and expression patterns of mlncRNAs.

Genetic Programming is a method to automatically create computer programs based on the principles of evolution. The problem of deceptiveness caused by complex dependencies among components of programs is challenging. It is important because it can misguide Genetic Programming to create suboptimal programs. Besides, a minor modification in the programs may lead to a notable change in the program behaviours and affect the final outputs. This article presents Grammar-Based Genetic Programming with Bayesian Classifiers (GBGPBC) in which the probabilistic dependencies among components of programs are captured using a set of Bayesian network classifiers. Our system was evaluated using a set of benchmark problems (the deceptive maximum problems, the royal tree problems, and the bipolar asymmetric royal tree problems). It was shown to be often more robust and more efficient in searching the best programs than other related Genetic Programming approaches in terms of the total number of fitness evaluation. We studied what factors affect the performance of GBGPBC and discovered that robust variants of GBGPBC were consistently weakly correlated with some complexity measures. Furthermore, our approach has been applied to learn a ranking program on a set of customers in direct marketing. Our suggested solutions help companies to earn significantly more when compared with other solutions produced by several well-known machine learning algorithms, such as neural networks, logistic regression, and Bayesian networks.

We consider a problem of stochastic online learning with general probabilistic graph feedback, where each directed edge in the feedback graph has probability pij. Two cases are covered. (a) The one-step case, where after playing arm i the learner observes a sample reward feedback of arm j with independent probability pij. (b) The cascade case where after playing arm i the learner observes feedback of all arms j in a probabilistic cascade starting from i – for each (i,j) with probability pij, if arm i is played or observed, then a reward sample of arm j would be observed with independent probability pij. Previous works mainly focus on deterministic graphs which corresponds to one-step case with pij ∈ {0,1}, an adversarial sequence of graphs with certain topology guarantees, or a specific type of random graphs. We analyze the asymptotic lower bounds and design algorithms in both cases. The regret upper bounds of the algorithms match the lower bounds with high probability.

With the rapid development of society and the economy and the increasing seriousness of environmental problems, renewable energy and high-quality energy services in low-carbon communities have become popular research topics. However, a large number of volatile distributed generation power systems in the community are connected to the grid. It is difficult to stabilize and efficiently interact with fragmented and isolated energy management systems, and it is difficult to meet energy management needs in terms of low-carbon emissions, stability, and intelligence. Therefore, by considering operation costs, pollution control costs, energy stability, and plug-in hybrid electric vehicles, this article proposes a regional energy supply model called community energy Internet and builds a low-carbon community energy adaptive management model for smart services. Then, to address energy supply instability, an adaptive feedback control mechanism developed based on model predictive control is introduced to adapt to the changing environment. Finally, a long short-term memory-recurrent neural network-based Tabu search is introduced to prevent the multiobjective particle swarm optimization algorithm from easily falling into a local optimum. The simulation results show that the proposed model can effectively realize the optimal allocation of energy, which solves the problem of fragmented energy islands caused by distributed power access. This method has quality of service benefits for users, such as cost, time, and stability, and realizes wide interconnections, high intelligence, and low-carbon efficiency of community energy management.

We generalize the setting of online clustering of bandits by allowing non-uniform distribution over user frequencies. A more efficient algorithm is proposed with simple set structures to represent clusters. We prove a regret bound for the new algorithm which is free of the minimal frequency over users. The experiments on both synthetic and real datasets consistently show the advantage of the new algorithm over existing methods.

MOTIVATION: Studies have shown that the accuracy of random forest (RF)-based scoring functions (SFs), such as RF-Score-v3, increases with more training samples, whereas that of classical SFs, such as X-Score, does not. Nevertheless, the impact of the similarity between training and test samples on this matter has not been studied in a systematic manner. It is therefore unclear how these SFs would perform when only trained on protein-ligand complexes that are highly dissimilar or highly similar to the test set. It is also unclear whether SFs based on machine learning algorithms other than RF can also improve accuracy with increasing training set size and to what extent they learn from dissimilar or similar training complexes. RESULTS: We present a systematic study to investigate how the accuracy of classical and machine-learning SFs varies with protein-ligand complex similarities between training and test sets. We considered three types of similarity metrics, based on the comparison of either protein structures, protein sequences or ligand structures. Regardless of the similarity metric, we found that incorporating a larger proportion of similar complexes to the training set did not make classical SFs more accurate. In contrast, RF-Score-v3 was able to outperform X-Score even when trained on just 32% of the most dissimilar complexes, showing that its superior performance owes considerably to learning from dissimilar training complexes to those in the test set. In addition, we generated the first SF employing Extreme Gradient Boosting (XGBoost), XGB-Score, and observed that it also improves with training set size while outperforming the rest of SFs. Given the continuous growth of training datasets, the development of machine-learning SFs has become very appealing. AVAILABILITY AND IMPLEMENTATION: https://github.com/HongjianLi/MLSF. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: Clustering molecular network is a typical method in system biology, which is effective in predicting protein complexes or functional modules. However, few studies have realized that biological molecules are spatial-temporally regulated to form a dynamic cellular network and only a subset of interactions take place at the same location in cells. RESULTS: In this study, considering the subcellular localization of proteins, we first construct a co-localization human protein interaction network (PIN) and systematically investigate the relationship between subcellular localization and biological functions. After that, we propose a Locational and Topological Overlap Model (LTOM) to preprocess the co-localization PIN to identify functional modules. LTOM requires the topological overlaps, the common partners shared by two proteins, to be annotated in the same localization as the two proteins. We observed the model has better correspondence with the reference protein complexes and shows more relevance to cancers based on both human and yeast datasets and two clustering algorithms, ClusterONE and MCL. CONCLUSION: Taking into consideration of protein localization and topological overlap can improve the performance of module detection from protein interaction networks.

Understanding the phenotypic drug response on cancer cell lines plays a vital role in anti-cancer drug discovery and re-purposing. The Genomics of Drug Sensitivity in Cancer (GDSC) database provides open data for researchers in phenotypic screening to build and test their models. Previously, most research in these areas starts from the molecular fingerprints or physiochemical features of drugs, instead of their structures. In this paper, a model called twin Convolutional Neural Network for drugs in SMILES format (tCNNS) is introduced for phenotypic screening. tCNNS uses a convolutional network to extract features for drugs from their simplified molecular input line entry specification (SMILES) format and uses another convolutional network to extract features for cancer cell lines from the genetic feature vectors respectively. After that, a fully connected network is used to predict the interaction between the drugs and the cancer cell lines. When the training set and the testing set are divided based on the interaction pairs between drugs and cell lines, tCNNS achieves 0.826, 0.831 for the mean and top quartile of the coefficient of determinant ( R 2 ) respectively and 0.909, 0.912 for the mean and top quartile of the Pearson correlation ( R p ) respectively, which are significantly better than those of the previous works (Ammad-Ud-Din et al., J Chem Inf Model 54:2347–9, 2014), (Haider et al., PLoS ONE 10:0144490, 2015), (Menden et al., PLoS ONE 8:61318, 2013). However, when the training set and the testing set are divided exclusively based on drugs or cell lines, the performance of tCNNS decreases significantly and R p and R 2 drop to barely above 0. Our approach is able to predict the drug effects on cancer cell lines with high accuracy, and its performance remains stable with less but high-quality data, and with fewer features for the cancer cell lines. tCNNS can also solve the problem of outliers in other feature space. Besides achieving high scores in these statistical metrics, tCNNS also provides some insights into the phenotypic screening. However, the performance of tCNNS drops in the blind test.

BACKGROUND: Development of new drugs is a time-consuming and costly process, and the cost is still increasing in recent years. However, the number of drugs approved by FDA every year per dollar spent on development is declining. Drug repositioning, which aims to find new use of existing drugs, attracts attention of pharmaceutical researchers due to its high efficiency. A variety of computational methods for drug repositioning have been proposed based on machine learning approaches, network-based approaches, matrix decomposition approaches, etc. RESULTS: We propose a novel computational method for drug repositioning. We construct and decompose three-dimensional tensors, which consist of the associations among drugs, targets and diseases, to derive latent factors reflecting the functional patterns of the three kinds of entities. The proposed method outperforms several baseline methods in recovering missing associations. Most of the top predictions are validated by literature search and computational docking. Latent factors are used to cluster the drugs, targets and diseases into functional groups. Topological Data Analysis (TDA) is applied to investigate the properties of the clusters. We find that the latent factors are able to capture the functional patterns and underlying molecular mechanisms of drugs, targets and diseases. In addition, we focus on repurposing drugs for cancer and discover not only new therapeutic use but also adverse effects of the drugs. In the in-depth study of associations among the clusters of drugs, targets and cancer subtypes, we find there exist strong associations between particular clusters. CONCLUSIONS: The proposed method is able to recover missing associations, discover new predictions and uncover functional clusters of drugs, targets and diseases. The clustering of drugs, targets and diseases, as well as the associations among the clusters, provides a new guiding framework for drug repositioning.

Drug discovery is an expensive and time consuming process, especially in the era of new technology, such as personalized medicine where tremendous experiments and analysis are needed before bringing new drugs to the market. While In vivo and In vitro experiments are expensive, In silico methods become important and they can reduce the cost in drug discovery by prioritizing the experiments in more efficient ways. In this paper, we propose a new convolution neural network based active learning model which helps to reduce the number of experiments needed in drug discovery. Using the drugs performance on other cell lines as assisting information, our model can precisely select the most promising drug from those candidates for a new cell line. Our model uses a deep neural network structure where there are two CNN channels for drugs and cell lines respectively, which are followed by a fulled connected network. The experimental results show that our model can achieve significant better performance than the existing methods.

Motivation: Moonlighting proteins are a class of proteins having multiple distinct functions, which play essential roles in a variety of cellular and enzymatic functioning systems. Although there have long been calls for computational algorithms for the identification of moonlighting proteins, research on approaches to identify moonlighting long non-coding RNAs (lncRNAs) has never been undertaken. Here, we introduce a novel methodology, MoonFinder, for the identification of moonlighting lncRNAs. MoonFinder is a statistical algorithm identifying moonlighting lncRNAs without a priori knowledge through the integration of protein interactome, RNA-protein interactions and functional annotation of proteins. Results: We identify 155 moonlighting lncRNA candidates and uncover that they are a distinct class of lncRNAs characterized by specific sequence and cellular localization features. The non-coding genes that transcript moonlighting lncRNAs tend to have shorter but more exons and the moonlighting lncRNAs have a variable localization pattern with a high chance of residing in the cytoplasmic compartment in comparison to the other lncRNAs. Moreover, moonlighting lncRNAs and moonlighting proteins are rather mutually exclusive in terms of both their direct interactions and interacting partners. Our results also shed light on how the moonlighting candidates and their interacting proteins implicated in the formation and development of cancers and other diseases. Availability and implementation: The code implementing MoonFinder is supplied as an R package in the supplementary material. Supplementary information: Supplementary data are available at Bioinformatics online.

To efficiently and effectively monitor and mitigate air pollution in the urban environment, it is of paramount importance to integrate into a unified whole air pollutant concentration databases coming from different sources including the ground-based stations, mobile sensors, remote sensing, atmospheric-chemical-transport models and social media for the analysis and unraveling of the complex air pollution processes in space and time. This study constructs and implements for the first time a prototype of the fully integrated air pollution decision support system (APDSS) that put together in an integrated manner all relevant multi-scale, multi-type and multi-source data for decision-making on urban air pollution. The prototype contains the main system that handles the multi-source, multi-type and multi-scale databases, queries, visualization and data mining algorithms and the integrated modules that individually and holistically capitalize on the power of the ground-based stations, ground and aerial mobile sensors, satellite-borne remote-sensing technologies, atmospheric-chemical-transport models and social media. It renders a solid scientific foundation and system development methodology for the study of the spatiotemporal air pollution profiles crucial to the mitigation of urban air pollution. Real-life applications of the prototype are employed to illustrate the functionality of the APDSS.

Rapid economic development has been observed worldwide, which has caused environmental problems to worsen. Thus, the Energy Internet (EI), which accesses renewable energy and provides high-quality power services, has recently become a hot issue. As a subnet of the EI, an energy local area network (ELAN) consists of renewable power generation equipment, controllable distributed power generation equipment, storage systems, electric vehicles, and a large number of loads. Energy management is required for economic, environmental, and safety considerations. This paper proposes an energy management optimization model that addresses ELAN operations and includes pollution treatment fees; this model provides intelligent control of the charging and discharging of plug-in hybrid electric vehicles (PHEVs). This model achieves a nonlinear energy management optimization for an ELAN. To promote optimal performance, an improved comprehensive learning particle swarm optimization (CLPSO) algorithm is presented; it combines Tabu Search (TS) and CLPSO to avoid local optima. To verify the performance of our model, two experimental scenarios are built. The simulation results show that our energy management optimization model fulfills the optimal allocation of energy and that the PHEV intelligent charging/discharging strategy promotes economic benefits for the network.

Understanding binding cores is of fundamental importance in deciphering Protein-DNA (TF-TFBS) binding and gene regulation. Limited by expensive experiments, it is promising to discover them with variations directly from sequence data. Although existing computational methods have produced satisfactory results, they are one-to-one mappings with no site-specific information on residue/nucleotide variations, where these variations in binding cores may impact binding specificity. This study presents a new representation for modeling binding cores by incorporating variations and an algorithm to discover them from only sequence data. Our algorithm takes protein and DNA sequences from TRANSFAC (a Protein-DNA Binding Database) as input; discovers from both sets of sequences conserved regions in Aligned Pattern Clusters (APCs); associates them as Protein-DNA Co-Occurring APCs; ranks the Protein-DNA Co-Occurring APCs according to their co-occurrence, and among the top ones, finds three-dimensional structures to support each binding core candidate. If successful, candidates are verified as binding cores. Otherwise, homology modeling is applied to their close matches in PDB to attain new chemically feasible binding cores. Our algorithm obtains binding cores with higher precision and much faster runtime ( ≥ 1,600x) than that of its contemporaries, discovering candidates that do not co-occur as one-to-one associated patterns in the raw data. AVAILABILITY: http://www.pami.uwaterloo.ca/~ealee/files/tcbbPnDna2015/Release.zip.

BACKGROUND: Pose generation error is usually quantified as the difference between the geometry of the pose generated by the docking software and that of the same molecule co-crystallised with the considered protein. Surprisingly, the impact of this error on binding affinity prediction is yet to be systematically analysed across diverse protein-ligand complexes. RESULTS: Against commonly-held views, we have found that pose generation error has generally a small impact on the accuracy of binding affinity prediction. This is also true for large pose generation errors and it is not only observed with machine-learning scoring functions, but also with classical scoring functions such as AutoDock Vina. Furthermore, we propose a procedure to correct a substantial part of this error which consists of calibrating the scoring functions with re-docked, rather than co-crystallised, poses. In this way, the relationship between Vina-generated protein-ligand poses and their binding affinities is directly learned. As a result, test set performance after this error-correcting procedure is much closer to that of predicting the binding affinity in the absence of pose generation error (i.e. on crystal structures). We evaluated several strategies, obtaining better results for those using a single docked pose per ligand than those using multiple docked poses per ligand. CONCLUSIONS: Binding affinity prediction is often carried out on the docked pose of a known binder rather than its co-crystallised pose. Our results suggest than pose generation error is in general far less damaging for binding affinity prediction than it is currently believed. Another contribution of our study is the proposal of a procedure that largely corrects for this error. The resulting machine-learning scoring function is freely available at http://istar.cse.cuhk.edu.hk/rf-score-4.tgz and http://ballester.marseille.inserm.fr/rf-score-4.tgz .

Gene regulatory network (GRN), which refers to the complex interactions with time delays between TFs and other genes, plays an important role in the working of the cell. Therefore inferring the GRN is crucial to studying diseases related to malfunctioning of the cell. Even with high-throughput technology, time series expression data is still limited compared to the network size, which poses significant challenge to inferring large GRN. Since GRNs are known to be modular, or hierarchically modular, we propose to exploit this by first inferring an initial GRN using CLINDE, then decomposing it into possibly overlapping subnetworks, then re-learning the subnetworks using either CLINDE or DD-lasso, and lastly merging the subnetworks. We have performed extensive experiments on synthetic data to test this strategy on both modular and hierarchically modular networks with 500 and 1000 genes, using either a long time series or several short time series. Results show that the strategy does improve GRN inference with statistical significance. Also, the algorithm is robust to different variance and slight deviation of Gaussianity for the error terms.

Classification on pseudoknots existence is a challenging and meaningful problem in Bioinformatics. As predicting RNA secondary structures with pseudoknots is NP-complete problem while predicting pseudoknot-free structures can be done in O(n3) time, if a preliminary pseudoknots existence classification of RNA sequence can be done before the prediction, the classification result can enhance the efficiency of RNA secondary structure prediction. In this paper, a classification of the existence of pseudoknots in an RNA sequence is presented. A set of features have been chosen by partial sequence content and thousands of RNA sequences with validated structures are used to train the classifier. Using a validated testing dataset, this classification method is shown to achieve a very good performance that the best result get 87% accuracy in 10-fold cross validation and around 75% accuracy in testing data. Moreover it may reveal how partial sequence content can affect the formation of pseudoknots.

BACKGROUND: Inferring gene regulatory network (GRN) has been an important topic in Bioinformatics. Many computational methods infer the GRN from high-throughput expression data. Due to the presence of time delays in the regulatory relationships, High-Order Dynamic Bayesian Network (HO-DBN) is a good model of GRN. However, previous GRN inference methods assume causal sufficiency, i.e. no unobserved common cause. This assumption is convenient but unrealistic, because it is possible that relevant factors have not even been conceived of and therefore un-measured. Therefore an inference method that also handles hidden common cause(s) is highly desirable. Also, previous methods for discovering hidden common causes either do not handle multi-step time delays or restrict that the parents of hidden common causes are not observed genes. RESULTS: We have developed a discrete HO-DBN learning algorithm that can infer also hidden common cause(s) from discrete time series expression data, with some assumptions on the conditional distribution, but is less restrictive than previous methods. We assume that each hidden variable has only observed variables as children and parents, with at least two children and possibly no parents. We also make the simplifying assumption that children of hidden variable(s) are not linked to each other. Moreover, our proposed algorithm can also utilize multiple short time series (not necessarily of the same length), as long time series are difficult to obtain. CONCLUSIONS: We have performed extensive experiments using synthetic data on GRNs of size up to 100, with up to 10 hidden nodes. Experiment results show that our proposed algorithm can recover the causal GRNs adequately given the incomplete data. Using the limited real expression data and small subnetworks of the YEASTRACT network, we have also demonstrated the potential of our algorithm on real data, though more time series expression data is needed.

Inferring gene regulatory network (GRN) from the microarray expression data is an important problem in Bioinformatics, because knowing the GRN is an essential first step in understanding the inner workings of the cell and the related diseases. Time delays exist in the regulatory effects from one gene to another due to the time needed for transcription, translation, and to accumulate a sufficient number of needed proteins. Also, it is known that the delays are important for oscillatory phenomenon. Therefore, it is crucial to develop a causal gene network model, preferably as a function of time. In this paper, we propose an algorithm CLINDE to infer causal directed links in GRN with time delays and regulatory effects in the links from time-series microarray gene expression data. It is one of the most comprehensive in terms of features compared to the state-of-the-art discrete gene network models. We have tested CLINDE on synthetic data, the in vivo IRMA (On and Off) datasets and the [1] yeast expression data validated using KEGG pathways. Results show that CLINDE can effectively recover the links, the time delays and the regulatory effects in the synthetic data, and outperforms other algorithms in the IRMA in vivo datasets.

Understanding binding cores is of fundamental importance in deciphering Protein-DNA (TF-TFBS) binding and for the deep understanding of gene regulation. Traditionally, binding cores are identified in resolved high-resolution 3D structures. However, it is expensive, labor-intensive and time-consuming to obtain these structures. Hence, it is promising to discover binding cores computationally on a large scale. Previous studies successfully applied association rule mining to discover binding cores from TF-TFBS binding sequence data only. Despite the successful results, there are limitations such as the use of tight support and confidence thresholds, the distortion by statistical bias in counting pattern occurrences, and the lack of a unified scheme to rank TF-TFBS associated patterns. In this study, we proposed an association rule mining algorithm incorporating statistical measures and ranking to address these limitations. Experimental results demonstrated that, even when the threshold on support was lowered to one-tenth of the value used in previous studies, a satisfactory verification ratio was consistently observed under different confidence levels. Moreover, we proposed a novel ranking scheme for TF-TFBS associated patterns based on p-values and co-support values. By comparing with other discovery approaches, the effectiveness of our algorithm was demonstrated. Eighty-four binding cores with PDB support are uniquely identified.

Understanding binding cores is of fundamental importance in deciphering Protein-DNA (TF-TFBS) binding and gene regulation. Variations (or mutations) in binding cores are ubiquitous and have different levels of effects on the binding specificity. To alleviate expensive experiments, we have developed a new method to discover directly from sequence data binding cores and study the effect due to variations. Although existing computational methods have produced satisfactory TF-TFBS binding cores, they are only one-to-one mappings with no site-specific information on residue/nucleotide variations; and also are largely overlapped. In this study, we propose a new representation for modeling TF-TFBS binding with variants known as TF-TFBS Co-Supportive Aligned Pattern Clusters (APCs), which are more compact, with more details for site-specific variants, and biologically more intuitive for analysis. To achieve this task, we have also developed an algorithm to discover TF-TFBS Co-Supportive APCs to capture binding cores at a higher precision with much faster runtime (≥1600X) comparing to other methods. The variants in TF-TFBS Co-Supportive APCs are also statistically analyzed and demonstrated that they can assist homology modeling to synthesize new biological knowledge.

Grammar-Based Genetic Programming (GBGP) improves the search performance of Genetic Programming (GP) by formalizing constraints and domain specific knowledge in grammar. The building blocks (i.e. the functions and the terminals) in a program can be dependent. Random crossover and mutation destroy the dependence with a high probability, hence breeding a poor program from good programs. Understanding on the syntactic and semantic in the grammar plays an important role to boost the efficiency of GP by reducing the number of poor breeding. Therefore, approaches have been proposed by introducing context sensitive ingredients encoded in probabilistic models. In this paper, we propose Grammar-Based Genetic Programming with Bayesian Network (BGBGP) which learns the dependence by attaching a Bayesian network to each derivation rule and demonstrates its effectiveness in two benchmark problems.

Grammar-Based Genetic Programming formalizes constraints on the solution structure based on domain knowledge to reduce the search space and generate grammatically correct individuals. Nevertheless, building blocks in a program can often be dependent, so the effective search space can be further reduced. Approaches have been proposed to learn the dependence using probabilistic models and shown to be useful in finding the optimal solutions with complex structure. It raises questions on how to use the individuals in the population to uncover the underlying dependence. Usually, only the good individuals are selected. To model the dependence better, we introduce Grammar-Based Genetic Programming with Bayesian Network Classifier (GBGPBC) which also uses poorer individuals. With the introduction of class labels, we further propose a refinement technique on probability distribution based on class label. Our results show that GBGPBC performs well on two benchmark problems. These techniques boost the performance of our system.

We present iSyn, a WebGL-based tool for interactivede novo drug design. It features an evolutionary algorithm that automatically designs novel ligands with drug-like properties and synthetic feasibility using click chemistry. Isyn interfaces with our popular and fast molecular docking engine idock, remarkably reducing the evaluation and ranking time of drug candidates. Furthermore, inspired by our user friendly and high-performance WebGL visualizer iview, our iSyn also implements a tailor-made interactive visualizer to aid novel drug design. We believe iSyn can supplement the efforts of medicinal chemists in drug discovery research. To illustrate the utility of iSyn in generating novelligands ex nihilo, we designed predicted inhibitors of two important drug targets, which are RNA editing ligase 1(REL1) from T. Brucei, the etiological agent of African sleeping sickness, and cyclin-dependent kinase 2 (CDK2), a positive regulator of eukaryotic cell cycle progression. Results show that iSyn managed to significantly enhance the predicted binding affinity of the best generated ligand by more than 3 orders of magnitude in potency. Isyn is written in C++, Python, HTML5 and JavaScript. It is free and open source, available athttp://istar.cse.cuhk.edu.hk/iSyn.tgz. It has been tested successfully on both Linux and Windows.

BACKGROUND: Visualization of protein-ligand complex plays an important role in elaborating protein-ligand interactions and aiding novel drug design. Most existing web visualizers either rely on slow software rendering, or lack virtual reality support. The vital feature of macromolecular surface construction is also unavailable. RESULTS: We have developed iview, an easy-to-use interactive WebGL visualizer of protein-ligand complex. It exploits hardware acceleration rather than software rendering. It features three special effects in virtual reality settings, namely anaglyph, parallax barrier and oculus rift, resulting in visually appealing identification of intermolecular interactions. It supports four surface representations including Van der Waals surface, solvent excluded surface, solvent accessible surface and molecular surface. Moreover, based on the feature-rich version of iview, we have also developed a neat and tailor-made version specifically for our istar web platform for protein-ligand docking purpose. This demonstrates the excellent portability of iview. CONCLUSIONS: Using innovative 3D techniques, we provide a user friendly visualizer that is not intended to compete with professional visualizers, but to enable easy accessibility and platform independence.

BACKGROUND: State-of-the-art protein-ligand docking methods are generally limited by the traditionally low accuracy of their scoring functions, which are used to predict binding affinity and thus vital for discriminating between active and inactive compounds. Despite intensive research over the years, classical scoring functions have reached a plateau in their predictive performance. These assume a predetermined additive functional form for some sophisticated numerical features, and use standard multivariate linear regression (MLR) on experimental data to derive the coefficients. RESULTS: In this study we show that such a simple functional form is detrimental for the prediction performance of a scoring function, and replacing linear regression by machine learning techniques like random forest (RF) can improve prediction performance. We investigate the conditions of applying RF under various contexts and find that given sufficient training samples RF manages to comprehensively capture the non-linearity between structural features and measured binding affinities. Incorporating more structural features and training with more samples can both boost RF performance. In addition, we analyze the importance of structural features to binding affinity prediction using the RF variable importance tool. Lastly, we use Cyscore, a top performing empirical scoring function, as a baseline for comparison study. CONCLUSIONS: Machine-learning scoring functions are fundamentally different from classical scoring functions because the former circumvents the fixed functional form relating structural features with binding affinities. RF, but not MLR, can effectively exploit more structural features and more training samples, leading to higher prediction performance. The future availability of more X-ray crystal structures will further widen the performance gap between RF-based and MLR-based scoring functions. This further stresses the importance of substituting RF for MLR in scoring function development.

RNAs are functionally important in many biological processes. Predicting secondary structures of RNAs can help understanding 3D structures and functions of RNAs. However, RNA secondary structure prediction with pseudoknots is NP-complete. Predicting whether the RNAs contain pseudoknots in advance can save computation time as secondary structure prediction without pseudoknots is much faster. In this paper, we use k-mer occurrences as attributes to predict whether the RNAs have pseudoknots in the secondary structure. The results show two classifiers can predict 90% of the instance correctly.

The free energy (evaluation) models used in RNA secondary structure prediction are one of the most important reasons that makes the prediction a challenging computational problem in Bioinformatics. These models are the key factor determining the accuracy of the prediction algorithms. Previously we have developed a method called GAknot that has obtained good performance on predicting RNA secondary structures with pseudoknots. In this paper, we propose a new free energy model. We first select a number of RNA sequences from a database which contains known RNA secondary structures as a training dataset for learning this new model. From the training dataset, we then extract base pairs patterns in subsequences of pairs of k-mers from the stems of each sequence in the training data and use the patterns to formulate penalty factors. We modify the energy model by adding these penalty factors. Combined with the new modified energy model, the prediction performance of GAknot has been improved significantly. GAknot with the new modified energy model is shown to be the best method in comparison with two state-of-the-art algorithms using a commonly used testing dataset. The penalty factors of the new energy model and dataset can be downloaded at http://appsrv.cse.cuhk.edu.hk/~kktong/NewModel.

DNA motif discovery is an important problem for deciphering protein-DNA bindings in gene regulation. To discover generic spaced motifs which have multiple conserved patterns separated by wild-cards called spacers, the genetic algorithm (GA) based GASMEN has been proposed and shown to outperform related methods. However, the over-generic modeling of any number of spacers increases the optimization difficulty in practice. In protein-DNA binding case studies, complicated spaced motifs are rare while dimers with single spacers are more common spaced motifs. Moreover, errors (mismatches) in a conserved pattern are not arbitrarily distributed as certain highly conserved nucleotides are essential to maintain bindings. Motivated by better optimization in real applications, we have developed a new method, which is GA for Dimer-led and Error-restricted Spaced Motifs (GADESM). Common spaced motifs are paid special attention to using dimer-led initialization in the population initialization. The results on real datasets show that the dimer-led initialization in GADESM achieves better fitness than GASMEN with statistical significance. With additional error-restricted motif occurrence retrieval, GADESM has shown better performance than GASMEN on both comprehensive simulation data and a real ChIP-seq case study.

Non-coding RNAs are RNA molecules that do not translate into proteins. These RNAs are functional important in many biological processes. Their biological functions are highly related to their interaction partners. RNA-RNA interactions are one of the possibilities. It is desired to use computational methods to study and predict the interaction partners of non-coding RNAs. Most recent programs for RNA-RNA interaction prediction programs are based on Turner Energy Model. Some papers show that the free energy of RNA-structures is lower than random sequences but it is not statistically significant. It shows that we may need to modify the energy model for different RNA structures applications. In this paper, we first study the RNA-RNA interaction pattern using experimental validated RNA-RNA interaction data, which are extracted from sRNATarBase. We study the sRNA-mRNA interaction data and extract some features of the RNA-RNA interaction patterns. Then we combine these features about interaction sites into the Turner Energy Model. We develop a genetic algorithm based program RIPGA to solve the RNA-RNA interaction prediction problem. We use genetic algorithm because the RNA-RNA interaction prediction is NP-hard and we are interested to find out good suboptimal solutions. We use an sRNA-mRNA interaction dataset to evaluate the performance of the modified energy model and compare the results with two state-of-the-art programs. The comparison of the original model and the modified model shows that the modified energy model has better performance in both sensitivity and positive predictive value (PPV). Comparing RIPGA with state-of-the-art programs, RIPGA have better sensitivity and comparable positive predictive value.

Predicting RNA secondary structure is a significant challenge in Bioinformatics especially including pseudoknots. There are so many researches proposed that pseudoknots have their own biological functions inside human body, so it is important to predict this kind of RNA secondary structures. There are several methods to predict RNA secondary structure, and the most common one is using minimum free energy. However, finding the minimum free energy to predict secondary structure with pseudoknots has been proven to be an NP-complete problem, so there are many heuristic approaches trying to solve this kind of problems. In this paper, we propose GAknot, a computational method using genetic algorithm (GA), to predict RNA secondary structure with pseudoknots. GAknot first generates a set of maximal stems, and then it tries to generate several individuals by different combinations of stems. After halting condition is reached, GAknot will output the best solution as the output of predicted secondary structure. By using two commonly used validation data sets, GAknot is shown to be a better prediction method in terms of accuracy and speed comparing to several competitive prediction methods. Source code and datasets can be downloaded.

BACKGROUND: Microarray technology is widely used in cancer diagnosis. Successfully identifying gene biomarkers will significantly help to classify different cancer types and improve the prediction accuracy. The regularization approach is one of the effective methods for gene selection in microarray data, which generally contain a large number of genes and have a small number of samples. In recent years, various approaches have been developed for gene selection of microarray data. Generally, they are divided into three categories: filter, wrapper and embedded methods. Regularization methods are an important embedded technique and perform both continuous shrinkage and automatic gene selection simultaneously. Recently, there is growing interest in applying the regularization techniques in gene selection. The popular regularization technique is Lasso (L1), and many L1 type regularization terms have been proposed in the recent years. Theoretically, the Lq type regularization with the lower value of q would lead to better solutions with more sparsity. Moreover, the L1/2 regularization can be taken as a representative of Lq (0 <q < 1) regularizations and has been demonstrated many attractive properties. RESULTS: In this work, we investigate a sparse logistic regression with the L1/2 penalty for gene selection in cancer classification problems, and propose a coordinate descent algorithm with a new univariate half thresholding operator to solve the L1/2 penalized logistic regression. Experimental results on artificial and microarray data demonstrate the effectiveness of our proposed approach compared with other regularization methods. Especially, for 4 publicly available gene expression datasets, the L1/2 regularization method achieved its success using only about 2 to 14 predictors (genes), compared to about 6 to 38 genes for ordinary L1 and elastic net regularization approaches. CONCLUSIONS: From our evaluations, it is clear that the sparse logistic regression with the L1/2 penalty achieves higher classification accuracy than those of ordinary L1 and elastic net regularization approaches, while fewer but informative genes are selected. This is an important consideration for screening and diagnostic applications, where the goal is often to develop an accurate test using as few features as possible in order to control cost. Therefore, the sparse logistic regression with the L1/2 penalty is effective technique for gene selection in real classification problems.

Understanding protein-DNA interactions, specifically transcription factor (TF) and transcription factor binding site (TFBS) bindings, is crucial in deciphering gene regulation. The recent associated TF-TFBS pattern discovery combines one-sided motif discovery on both the TF and the TFBS sides. Using sequences only, it identifies the short protein-DNA binding cores available only in high-resolution 3D structures. The discovered patterns lead to promising subtype and disease analysis applications. While the related studies use either association rule mining or existing TFBS annotations, none has proposed any formal unified (both-sided) model to prioritize the top verifiable associated patterns. We propose the unified scores and develop an effective pipeline for associated TF-TFBS pattern discovery. Our stringent instance-level evaluations show that the patterns with the top unified scores match with the binding cores in 3D structures considerably better than the previous works, where up to 90 percent of the top 20 scored patterns are verified. We also introduce extended verification from literature surveys, where the high unified scores correspond to even higher verification percentage. The top scored patterns are confirmed to match the known WRKY binding cores with no available 3D structures and agree well with the top binding affinities of in vivo experiments.

AutoDock Vina is a competitive protein-ligand docking tool well known for its fast execution and high accuracy. Nevertheless, when docking a massive number of ligands, Vina has to be run multiple times, repeating receptor parsing and grid maps building over and over again. There are tremendous requests for revising Vina to reuse precalculated data and incorporate built-in support for virtual screening. Hence we developed idock, inheriting from AutoDock Vina the accurate scoring function and the efficient optimization algorithm, and significantly improving the fundamental implementation and numerical model for even faster execution. idock achieves a speedup of 3.3 in terms of CPU time and a speedup of 7.5 in terms of elapsed time on average. idock is free and open source, available at https://GitHub.com/HongjianLi/idock.

The studies of protein-DNA bindings between transcription factors (TFs) and transcription factor binding sites (TFBSs) are important bioinformatics topics. High-resolution (length;490) are shown promising in identifying accurate binding cores without using any 3D structures. While the current association rule mining method on this problem addresses exact sequences only, the most recent ad hoc method for approximation does not establish any formal model and is limited by experimentally known patterns. As biological mutations are common, it is desirable to formally extend the exact model into an approximate one. In this paper, we formalize the problem of mining approximate protein-DNA association rules from sequence data and propose a novel efficient algorithm to predict protein-DNA binding cores. Our two-phase algorithm first constructs two compact intermediate structures called frequent sequence tree (FS-Tree) and frequent sequence class tree (FSCTree). Approximate association rules are efficiently generated from the structures and bioinformatics concepts (position weight matrix and information content) are further employed to prune meaningless rules. Experimental results on real data show the performance and applicability of the proposed algorithm.

A web-based system is developed for the analysis and prediction of tropical cyclones, particularly their landfalls and recurvatures. To facilitate accessibility to the system, its development is based on Google Maps application programming interface (API), Java and client/server architecture. In addition to the construction of a powerful query system for the multi-source, multi-scale and multi-level tropical cyclone database, data mining approach and dynamic modelling approach have been implemented and integrated for effective and efficient analysis, prediction and visualization of tropical cyclone movements. The system can be accessed worldwide by researchers, professionals and the general public. It is thus a powerful system for research, real-life application and knowledge dissemination. Its extensibility and user-friendliness pave the road for further development and enable more in-depth analysis and real-time operation.

Nonlinear integrals (NIs) are useful integration tools. It can get a set of virtual values by projecting original data onto a virtual space for classification purpose using NIs. The classical NIs implement projection along a line with respect to the features. But, in many cases, the linear projection cannot achieve good performance for classification or regression due to the limitation of the integrand. The linear function used for the integrand is just a special type of function with respect to the features. In this paper, we propose a nonlinear integrals with polynomial kernel (NIPK). A polynomial function with respect to the features is used as the integrand of NIs. It enables the projection to be along different types of curves to the virtual space so that the virtual values gotten by NIs can be better regularized and have higher separation power for classification. We use genetic algorithm to learn the fuzzy measures so that a larger solution space can be searched. To test the capability of the NIPK, we apply it to classification on several benchmark datasets and a bioinformatics project. Experiments show that there is evident improvement on performance for the NIPK compared to classical NIs. © 2011 Wiley Periodicals, Inc.

A new nonlinear multiregression model based on a pair of extreme nonlinear integrals, upper and lower nonlinear integrals with respect to signed fuzzy measure, is established in this paper. A data set with the predictive features and the relevant objective feature is required for estimating the regression coefficients. Owing to the nonadditivity of the model, a multiobjective optimization using genetic algorithm is adopted to search for the optimized solution in the regression problem. Applying such a nonlinear multiregression model, an interval prediction for the value of the objective feature can be made once a new observation of predictive features is available. We apply our model on synthetic data and weather problem. The results testify the performance of the multiregression based on upper and lower nonlinear integrals. © 2012 Wiley Periodicals, Inc.

Identifying the unknown transcription factor binding sites (TFBSs) is a fundamental and important component for understanding gene regulation as well as life mechanisms. The corresponding de novo motif discovery problem in bioinformatics is formulated as pattern discovery from strings, where challenges come from both modeling and optimization, because the short TFBSs are weak signals in massive and noisy experimental data. While genetic algorithms have been widely applied to the problem, recent memetic algorithms (MAs) employing local operators demonstrate the superiority in both effectiveness and efficiency. In this paper, we propose and study various MA components including local operators and models for motif discovery, through the newly established MA framework. The demonstrated optimization and modeling capabilities are analyzed in-depth on real datasets and their noisy versions. Selected optimal MAs show significantly improved performance over state-of-the-art methods in extensive tests including the blind test on the eukaryotic benchmark. This paper serves as the first systematic study of MAs on de novo motif discovery, where important issues are highlighted in the analyses of MA design. The comprehensive component categorization and the MA framework provide a useful platform for future MA developments, especially on the newly emerging chromatin immunoprecipitation followed by sequencing data.

Discovering new drugs for emerging diseases has been a challenging task. There are numerous drug design techniques including fragment-based and diversity-oriented methods but their accuracies and efficiencies are low. By incorporating visualisation, biomedical experts can interact with the process to produce drug-like ligands more efficiently. The paper presents an interactive drug design algorithm which generates lead candidates against a protein. A set of drug candidates, created by an in house fragment-based method and docked on the target protein, are visualised in the virtual reality settings. Biomedical experts can investigate and select some of the ligands for further processing, aided with distance and bonding information. It also assists the user to drag and rotate the ligand to the binding site they find suitable. The algorithm runs iteratively and improves the quality of lead candidates every step. The paper compares the quality of resulting ligands between interactive and automatic approaches.

MOTIVATION: The bindings between transcription factors (TFs) and transcription factor binding sites (TFBSs) are fundamental protein-DNA interactions in transcriptional regulation. Extensive efforts have been made to better understand the protein-DNA interactions. Recent mining on exact TF-TFBS-associated sequence patterns (rules) has shown great potentials and achieved very promising results. However, exact rules cannot handle variations in real data, resulting in limited informative rules. In this article, we generalize the exact rules to approximate ones for both TFs and TFBSs, which are essential for biological variations. RESULTS: A progressive approach is proposed to address the approximation to alleviate the computational requirements. Firstly, similar TFBSs are grouped from the available TF-TFBS data (TRANSFAC database). Secondly, approximate and highly conserved binding cores are discovered from TF sequences corresponding to each TFBS group. A customized algorithm is developed for the specific objective. We discover the approximate TF-TFBS rules by associating the grouped TFBS consensuses and TF cores. The rules discovered are evaluated by matching (verifying with) the actual protein-DNA binding pairs from Protein Data Bank (PDB) 3D structures. The approximate results exhibit many more verified rules and up to 300% better verification ratios than the exact ones. The customized algorithm achieves over 73% better verification ratios than traditional methods. Approximate rules (64-79%) are shown statistically significant. Detailed variation analysis and conservation verification on NCBI records demonstrate that the approximate rules reveal both the flexible and specific protein-DNA interactions accurately. The approximate TF-TFBS rules discovered show great generalized capability of exploring more informative binding rules.

UNLABELLED: Sequencing reads generated by RNA-sequencing (RNA-seq) must first be mapped back to the genome through alignment before they can be further analyzed. Current fast and memory-saving short-read mappers could give us a quick view of the transcriptome. However, they are neither designed for reads that span across splice junctions nor for repetitive reads, which can be mapped to multiple locations in the genome (multi-reads). Here, we describe a new software package: ABMapper, which is specifically designed for exploring all putative locations of reads that are mapped to splice junctions or repetitive in nature. AVAILABILITY AND IMPLEMENTATION: The software is freely available at: http://abmapper.sourceforge.net/. The software is written in C++ and PERL. It runs on all major platforms and operating systems including Windows, Mac OS X and LINUX.

BACKGROUND: RNA sequencing (RNA-seq) measures gene expression levels and permits splicing analysis. Many existing aligners are capable of mapping millions of sequencing reads onto a reference genome. For reads that can be mapped to multiple positions along the reference genome (multireads), these aligners may either randomly assign them to a location, or discard them altogether. Either way could bias downstream analyses. Meanwhile, challenges remain in the alignment of reads spanning across splice junctions. Existing splicing-aware aligners that rely on the read-count method in identifying junction sites are inevitably affected by sequencing depths. RESULTS: The distance between aligned positions of paired-end (PE) reads or two parts of a spliced read is dependent on the experiment protocol and gene structures. We here proposed a new method that employs an empirical geometric-tail (GT) distribution of intron lengths to make a rational choice in multireads selection and splice-sites detection, according to the aligned distances from PE and sliced reads. CONCLUSIONS: GT models that combine sequence similarity from alignment, and together with the probability of length distribution, could accurately determine the location of both multireads and spliced reads.

Extraction of meaningful information from large experimental data sets is a key element in bioinformatics research. One of the challenges is to identify genomic markers in Hepatitis B Virus (HBV) that are associated with HCC (liver cancer) development by comparing the complete genomic sequences of HBV among patients with HCC and those without HCC. In this study, a data mining framework, which includes molecular evolution analysis, clustering, feature selection, classifier learning, and classification, is introduced. Our research group has collected HBV DNA sequences, either genotype B or C, from over 200 patients specifically for this project. In the molecular evolution analysis and clustering, three subgroups have been identified in genotype C and a clustering method has been developed to separate the subgroups. In the feature selection process, potential markers are selected based on Information Gain for further classifier learning. Then, meaningful rules are learned by our algorithm called the Rule Learning, which is based on Evolutionary Algorithm. Also, a new classification method by Nonlinear Integral has been developed. Good performance of this method comes from the use of the fuzzy measure and the relevant nonlinear integral. The nonadditivity of the fuzzy measure reflects the importance of the feature attributes as well as their interactions. These two classifiers give explicit information on the importance of the individual mutated sites and their interactions toward the classification (potential causes of liver cancer in our case). A thorough comparison study of these two methods with existing methods is detailed. For genotype B, genotype C subgroups C1, C2, and C3, important mutation markers (sites) have been found, respectively. These two classification methods have been applied to classify never-seen-before examples for validation. The results show that the classification methods have more than 70 percent accuracy and 80 percent sensitivity for most data sets, which are considered high as an initial scanning method for liver cancer diagnosis.

RNA sequencing (RNA-Seq) measures gene expression levels and permits splicing analysis. There are many aligners that provide ultra-fast mapping of millions of sequencing reads onto a reference genome. However, random assignment or removal of reads matching multiple positions along the reference genome could bias downstream analyses. Meanwhile, challenges remain in the alignment of reads spanning across splice junctions. Existing splicing-aware aligners that rely on the read-count method in identifying junction sites are inevitably affected by sequencing depths. We here proposed a new method that employs an empirical geometric-tail (GT) distribution of intron lengths to make a rational choice in multireads selection and junction sites detection.

In this work, we define generalized (sequence) patterns, which is based on several real Biological problems, including transcription factors (TFs) binding to transcription factor binding sites (TFBSs), cis-regulatory modules, protein domain analysis, and alternative splicing etc. Simply speaking, a generalized pattern is composed of several substrings with gaps in-between two substrings. We propose a generalized pattern matching algorithm that uses a complementary dualseeding strategy, which is sensitive to errors (both mismatches and indels). We also develop a generalized pattern matching tool, which is to our knowledge the first ever developed specially for generalized pattern matching. Rather than replacing the existing general purpose matching tools, such as BLAST, BLAT, and PatternHunter etc, our tool provides an alternative and helps users to solve real problems, especially those that can be modeled as generalized patterns. We use data randomly sampled from reference sequences of human genome (NCBI build v18) in experiments, and hit 98.74% generalized patterns on average. The tool runs on both LINUX and Windows platforms, and the memory peak goes to a little bit larger than 1GB only.

DNA motif discovery is an important problem for deciphering gene regulation. Motifs usually contain gaps (spaced) and are more complex than contiguously conserved (monad) patterns. Existing algorithms mostly address monad motifs, and methods for spaced motifs impose various constraints on gaps, which may affect the discovery of complex motifs. In this paper, we propose Genetic Algorithm (GA) for Spaced Motifs Elicitation on Nucleotides (GASMEN), which searches from a wide range of possible widths (4-25) and relaxes substantial constraints. GASMEN employs submotif indexing to partition the search space into smaller sub-space for GA to easier reach optimality. Multiple-motif control is employed and probabilistic refinements are proposed to improve motif quality respectively. The preliminary results on real spaced motifs demonstrate that GASMEN is promising to find more accurate motifs and optimal widths, compared with the state-of-the-art method, SPACE. GASMEN is also capable of finding monad motifs, outperforming both Weeder and SPACE on most of the 8 real datasets.

Motif discovery is an important Bioinformatics problem for deciphering gene regulation. Numerous sequence-based approaches have been proposed employing human specialist motif models (evaluation functions), but performance is so unsatisfactory on benchmarks that the underlying information seems to have already been exploited and have doomed. However, we have found that even a simple modified representation still achieves considerably high performance on a challenging benchmark, implying potential for sequence-based motif discovery. Thus we raise the problem of learning motif evaluation functions. We employ Genetic programming (GP) which has the potential to evolve human competitive models. We take advantage of the terminal set containing specialist-model-like components and have tried three fitness functions. Results exhibit both great challenges and potentials. No models learnt can perform universally well on the challenging benchmark, where one reason may be the data appropriateness for sequence-based motif discovery. However, when applied on different widely-tested datasets, the same models achieve comparable performance to existing approaches based on specialist models. The study calls for further novel GP to learn different levels of effective evaluation models from strict to loose ones on exploiting sequence information for motif discovery, namely quantitative functions, cardinal rankings, and learning feasibility classifications.

This paper considers the protein structure prediction problem as a multimodal optimization problem. In particular, de novo protein structure prediction problems on the 3D Hydrophobic-Polar (HP) lattice model are tackled by evolutionary algorithms using multimodal optimization techniques. In addition, a new mutation approach and performance metric are proposed for the problem. The experimental results indicate that the proposed algorithms are more effective than the state-of-the-arts algorithms, even though they are simple.

Finding Transcription Factor Binding Sites, i.e., motif discovery, is crucial for understanding the gene regulatory relationship. Motifs are weakly conserved and motif discovery is an NP-hard problem. We propose a new approach called Cluster Refinement Algorithm for Motif Discovery (CRMD). CRMD employs a flexible statistical motif model allowing a variable number of motifs and motif instances. CRMD first uses a novel entropy-based clustering to find complete and good starting candidate motifs from the DNA sequences. CRMD then employs an effective greedy refinement to search for optimal motifs from the candidate motifs. The refinement is fast, and it changes the number of motif instances based on the adaptive thresholds. The performance of CRMD is further enhanced if the problem has one occurrence of motif instance per sequence. Using an appropriate similarity test of motifs, CRMD is also able to find multiple motifs. CRMD has been tested extensively on synthetic and real data sets. The experimental results verify that CRMD usually outperforms four other state-of-the-art algorithms in terms of the qualities of the solutions with competitive computing time. It finds a good balance between finding true motif instances and screening false motif instances, and is robust on problems of various levels of difficulty.

This paper presents an evolutionary algorithm, which we call Evolutionary Algorithm with Species-specific Explosion (EASE), for multimodal optimization. EASE is built on the Species Conserving Genetic Algorithm (SCGA), and the design is improved in several ways. In particular, it not only identifies species seeds, but also exploits the species seeds to create multiple mutated copies in order to further converge to the respective optimum for each species. Experiments were conducted to compare EASE and SCGA on four benchmark functions. Cross-comparison with recent rival techniques on another five benchmark functions was also reported. The results reveal that EASE has a competitive edge over the other algorithms tested.

BACKGROUND: Identification of transcription factor binding sites (TFBSs) is a central problem in Bioinformatics on gene regulation. de novo motif discovery serves as a promising way to predict and better understand TFBSs for biological verifications. Real TFBSs of a motif may vary in their widths and their conservation degrees within a certain range. Deciding a single motif width by existing models may be biased and misleading. Additionally, multiple, possibly overlapping, candidate motifs are desired and necessary for biological verification in practice. However, current techniques either prohibit overlapping TFBSs or lack explicit control of different motifs. RESULTS: We propose a new generalized model to tackle the motif widths by considering and evaluating a width range of interest simultaneously, which should better address the width uncertainty. Moreover, a meta-convergence framework for genetic algorithms (GAs), is proposed to provide multiple overlapping optimal motifs simultaneously in an effective and flexible way. Users can easily specify the difference amongst expected motif kinds via similarity test. Incorporating Genetic Algorithm with Local Filtering (GALF) for searching, the new GALF-G (G for generalized) algorithm is proposed based on the generalized model and meta-convergence framework. CONCLUSION: GALF-G was tested extensively on over 970 synthetic, real and benchmark datasets, and is usually better than the state-of-the-art methods. The range model shows an increase in sensitivity compared with the single-width ones, while providing competitive precisions on the E. coli benchmark. Effectiveness can be maintained even using a very small population, exhibiting very competitive efficiency. In discovering multiple overlapping motifs in a real liver-specific dataset, GALF-G outperforms MEME by up to 73% in overall F-scores. GALF-G also helps to discover an additional motif which has probably not been annotated in the dataset. http://www.cse.cuhk.edu.hk/%7Etmchan/GALFG/

Aiming at Automated Guided Vehicle (AGV) dynamic model characteristics, a Variable Structure Control based on genetic algorithm (GA) and least square-support vector machine (LS-SVM) was designed. Parameters, predetermined by conventional reaching law, were regulated by LS-SVM online. It was shown that system shattering is eliminated. Simulation results indicated that this method possesses the advantages of higher precision, greater adaptability and robustness, as compared to the conventional Variable Structure Control methods.

The problem of Transcription Factor Binding Sites identification or motif discovery is to identify the motif binding sites in the cis-regulatory regions of DNA sequences. The biological experiments are expensive and the problem is NP-hard computationally. We have proposed Estimation of Distribution Algorithm for Motif Discovery (EDAMD). We use Bayesian analysis to derive the fitness function to measure the posterior probability of a set of motif instances, which can be used to handle a variable number of motif instances in the sequences. EDAMD adopts a Gaussian distribution to model the distribution of the sets of motif instances, which is capable of capturing the bivariate correlation among the positions of motif instances. When a new Position Frequency Matrix (PFM) is generated from the Gaussian distribution, a new set of motif instances is identified based on the PFM via the Greedy Refinement operation. At the end of a generation, the Gaussian distribution is updated with the sets of motif instances. Since Greedy Refinement assumes a single motif instance on a sequence, a Post Processing operation based on the fitness function is used to find more motif instances after the evolution. The experiments have verified that EDAMD is comparable to or better than GAME and GALF on the real problems tested in this paper.

This work proposes a novel algorithm to do approximate matching in a database consisting of multiple sequences. We apply Agrep algorithm in an indexing structure, the r-cut numerical substring array (r-NSA). The structure basically indexes all the substrings of length r. The advantage of using the r-NSA is two-fold: (1) The space requirement of the r-NSA is much smaller than that of the other existing indexing structures, such as the generalized suffix tree. (2) We propose an algorithm to apply Agrep in the r-NSA, in which the substrings are processed sequentially. Since the common substrings are processed only once, the cost of our algorithm is smaller than that of the full scanning search by Agrep. Consequently, the matching time of our algorithm is also reduced. We design experiments to validate and compare the performance of our algorithm against the full scanning search by Agrep. We define the speed-up of our algorithm as the time required by the full scanning search by Agrep over that of our algorithm. We use eight sets of real DNA sequences in our experiments, and the results show that our algorithm achieves significant speed-up. We also investigate the speed-up of difference data sets, and analyze their differences in detail.

As an aggregation tool in information fusion and data mining, the Choquet integral is generalized to allow the involved set function being fuzzy-valued. A calculation formula of such a Choquet integral is developed when the universal set is finite, such as the set of attributes in a database.

MOTIVATION: Identification of transcription factor binding sites (TFBSs) plays an important role in deciphering the mechanisms of gene regulation. Recently, GAME, a Genetic Algorithm (GA)-based approach with iterative post-processing, has shown superior performance in TFBS identification. However, the basic GA in GAME is not elaborately designed, and may be trapped in local optima in real problems. The feature operators are only applied in the post-processing, but the final performance heavily depends on the GA output. Hence, both effectiveness and efficiency of the overall algorithm can be improved by introducing more advanced representations and novel operators in the GA, as well as designing the post-processing in an adaptive way. RESULTS: We propose a novel framework GALF-P, consisting of Genetic Algorithm with Local Filtering (GALF) and adaptive post-processing techniques (-P), to achieve both effectiveness and efficiency for TFBS identification. GALF combines the position-led and consensus-led representations used separately in current GAs and employs a novel local filtering operator to get rid of false positives within an individual efficiently during the evolutionary process in the GA. Pre-selection is used to maintain diversity and avoid local optima. Post-processing with adaptive adding and removing is developed to handle general cases with arbitrary numbers of instances per sequence. GALF-P shows superior performance to GAME, MEME, BioProspector and BioOptimizer on synthetic datasets with difficult scenarios and real test datasets. GALF-P is also more robust and reliable when further compared with GAME, the current state-of-the-art approach. AVAILABILITY: http://www.cse.cuhk.edu.hk/~tmchan/GALFP/.

In genetic programming (GP), evolving tree nodes separately would reduce the huge solution space. However, tree nodes are highly interdependent with respect to their fitness. In this paper, we propose a new GP framework, namely, instruction-matrix (IM)-based GP (IMGP), to handle their interactions. IMGP maintains an IM to evolve tree nodes and subtrees separately. IMGP extracts program trees from an IM and updates the IM with the information of the extracted program trees. As the IM actually keeps most of the information of the schemata of GP and evolves the schemata directly, IMGP is effective and efficient. Our experimental results on benchmark problems have verified that IMGP is not only better than those of canonical GP in terms of the qualities of the solutions and the number of program evaluations, but they are also better than some of the related GP algorithms. IMGP can also be used to evolve programs for classification problems. The classifiers obtained have higher classification accuracies than four other GP classification algorithms on four benchmark classification problems. The testing errors are also comparable to or better than those obtained with well-known classifiers. Furthermore, an extended version, called condition matrix for rule learning, has been used successfully to handle multiclass classification problems.

In this paper, the original Choquet integral is generalized as a Fuzzified Choquet Integral with a Fuzzy-valued Integrand (FCIFI), which supports a fuzzy-valued integrand and an integration result. The calculation of the FCIFI is established on the Choquet integral with an interval-valued integrand (CIII). The definitions, properties, and calculation algorithms of the CIII and the FCIFI are discussed and proposed in this paper. As a specific application scheme, we designed a CIII regression model for the regression problems involving interval-valued data. This CIII regression model has a self-learning ability through a double genetic algorithm. Finally, a daily temperature predictor based on the CIII regression model is discussed, where a series of experiments is implemented to validate the performance of the predictor by real weather records from the Hong Kong Observatory.

Identification of Transcription Factor Binding Site (TFBS) motifs in multiple DNA upstream sequences is important in understanding the mechanism of gene regulation. This identification problem is challenging because such motifs are usually weakly conserved due to evolutionary variation. Exhaustive search is intractable for finding long motifs because the combinatorial growth of the search space is exponential, thus heuristic methods are preferred. In this paper, we propose the Genetic Algorithm with Local Filtering (GALF) to address the problem, which combines and utilizes both position-led and consensus-led representations in present GA approaches. While position-led representation provides flexibility to move around the search space, it is likely to contain some "false positive" sites within an individual. This problem can be overcome by our local filtering operator, which employs consensus-led representation, while it needs less computation than alignments used in conventional consensus-led approaches. Thus both efficiency and accuracy can be achieved. The experimental results on real biological data show that our method can identify TFBSs more accurately and efficiently than other methods including GA-based ones, and is able to deal with relaxed motif widths with superior correctness.

The advances in kernel-based learning necessitate the study on solving a large-scale non-sparse positive definite linear system. To provide a deterministic approach, recent researches focus on designing fast matrixvector multiplication techniques coupled with a conjugate gradient method. Instead of using the conjugate gradient method, our paper proposes to use a domain decomposition approach in solving such a linear system. Its convergence property and speed can be understood within von Neumann’s alternating projection framework. We will report significant and consistent improvements in convergence speed over the conjugate gradient method when the approach is applied to recent machine learning problems. 1.

Experimental results show that parallel programs can be evolved more easily than sequential programs in genetic parallel programming (GPP). GPP is a novel genetic programming paradigm which evolves parallel program solutions. With the rapid development of lookup-table-based (LUT-based) field programmable gate arrays (FPGAs), traditional circuit design and optimization techniques cannot fully exploit the LUTs in LUT-based FPGAs. Based on the GPP paradigm, we have developed a combinational logic circuit learning system, called GPP logic circuit synthesizer (GPPLCS), in which a multilogic-unit processor is used to evaluate LUT circuits. To show the effectiveness of the GPPLCS, we have performed a series of experiments to evolve combinational logic circuits with two- and four-input LUTs. In this paper, we present eleven multi-output Boolean problems and their evolved circuits. The results show that the GPPLCS can evolve more compact four-input LUT circuits than the well-known LUT-based FPGA synthesis algorithms.

As a fuzzification of the Choquet integral, the defuzzified choquet integral with fuzzy-valued integrand (DCIFI) takes a fuzzy-valued integrand and gives a crisp-valued integration result. In this paper, the DCIFI acts as a projection to project high-dimensional heterogeneous fuzzy data to one-dimensional crisp data to handle the classification problems involving different data forms, such as crisp data, interval values, fuzzy numbers, and linguistic variables, simultaneously. The nonadditivity of the signed fuzzy measure applied in the DCIFI can represent the interaction among the measurements of features towards the discrimination of classes. Values of the signed fuzzy measure in the DCIFI are considered to be unknown parameters which should be learned before the classifier is used to classify new data. We have implemented a genetic algorithm (GA)-based adaptive classifier-learning algorithm to optimally learn the signed fuzzy measure values and the classified boundaries simultaneously. The performance of our algorithm has been tested both on synthetic and real data. The experimental results are satisfactory and outperform those of existing methods, such as the fuzzy decision trees and the fuzzy-neuro networks.

This correspondence introduces a multidrug cancer chemotherapy model to simulate the possible response of the tumor cells under drug administration. We formulate the model as an optimal control problem. The algorithm in this correspondence optimizes the multidrug cancer chemotherapy schedule. The objective is to minimize the tumor size under a set of constraints. We combine the adaptive elitist genetic algorithm with a local search algorithm called iterative dynamic programming (IDP) to form a new memetic algorithm (MA-IDP) for solving the problem. MA-IDP has been shown to be very efficient in solving the multidrug scheduling optimization problem.

Multi-class problem is the class of problems having more than one classes in the data set. Bayesian Network (BN) is a well-known algorithm handling the multi-class problem and is applied to different areas. But BN cannot handle continuous values. In contrast, Genetic Programming (GP) can handle continuous values and produces classification rules. However, GP is possible to produce cyclic rules representing tautologic, in which are useless for inference and expert systems. Co-evolutionary Rule-chaining Genetic Programming (CRGP) is the first variant of GP handling the multi-class problem and produces acyclic classification rules [16]. It employs backward chaining inference to carry out classification based on the acquired acyclic rule set. It can handle multi-classes; it can avoid cyclic rules; it can handle input attributes with continuous values; and it can learn complex relationships among the attributes. In this paper, we propose a novel algorithm, the Chaining Genetic Programming (CGP) learning a set of acyclic rules and to produce better results than the CRGP's. The experimental results demonstrate that the proposed algorithm has the shorter learning process and can produce more accurate acyclic classification rules.

Based on the theoretical guidance and existing recommendations for designing efficient genetic representations, we investigate a novel genetic representation — a splicing/decomposable (S/D) binary encoding in this paper. The S/D binary representation can be spliced and decomposed to describe potential solutions of the problem with different precisions by different number of uniform-salient building blocks (BBs). According to the characteristics of the S/D binary representation, genetic and evolutionary algorithms (GEAs) can be applied from the high scaled to the low scaled BBs sequentially to avoid genetic drift and improve GEAs’ performance. Our theoretical and empirical investigations reveal that the S/D binary representation is more proper than other existing binary encodings for GEAs searching.

In this paper, we propose two new anticancer drug scheduling models with different toxicity clearances according to kinetics of enzyme-catalyzed chemical reactions. We also present a sophisticated automating drug scheduling approach based on evolutionary computation and computer modeling. To explore multiple efficient drug scheduling policies, we use a multimodal optimization algorithm — adaptive elitist-population based genetic algorithm (AEGA) to solve the models, and discuss the situation of multiple optimal solutions under different parameter settings. The simulation results obtained by the new models match well with the clinical treatment experience, and can provide much more drug scheduling policies for a doctor to choose depending on the particular conditions of the patients.

Classification rule is a useful model in data mining. Given variable values, rules classify data items into different classes. Different rule learning algorithms are proposed, like Genetic Algorithm (GA) and Genetic Programming (GP). Rules can also be extracted from Bayesian Network (BN) and decision trees. However, all of them have disadvantages and may fail to get the best results. Both of GA and GP cannot handle cooperation among rules and thus, the learnt rules are likely to have many overlappings, i. e. more than one rules classify the same data items and different rules have different predictions. The conflicts among the rules reduce their understandability and increase their usage difficulty for expert systems. In contrast, rules extracted from BN and decision trees have no overlapping in nature. But BN can handle discrete values only and cannot represent higher-order relationships among variables. Moreover, the search space for decision tree learning is huge and thus, it is difficult to reach the global optimum. In this paper, we propose to use Functional Dependency Network (FDN) and MDL Genetic Programming (MDLGP) to learn a set of non-overlapping classification rules [17]. The FDN is an extension of BN; it can handle all kind of values; it can represent higher-order relationships among variables; and its learning search space is smaller than decision trees’. The experimental results demonstrate that the proposed method can successfully discover the target rules, which have no overlapping and have the highest classification accuracies.

Genetic Parallel Programming (GPP) is a novel Genetic Programming paradigm. Based on the GPP paradigm and a local search operator - FlowMap, a logic circuit synthesizing system integrating GPP and FlowMap, a Hybridized GPP based Logic Circuit Synthesizer (HGPPLCS) is developed. To show the effectiveness of the proposed HGPPLCS, six combinational logic circuit problems are used for evaluations. Each problem is run for 50 times. Experimental results show that both the lookup table counts and the propagation gate delays of the circuits collected are better than those obtained by conventional design or evolved by GPP alone. For example, in a 6-bit one counter experiment, we obtained combinational digital circuits with 8 four-input lookup tables in 2 gate level on average. It utilizes 2 lookup tables and 3 gate levels less than circuits evolved by GPP alone.

In this paper, we introduce a new genetic representation --- a splicing/decomposable (S/D) binary encoding, which was proposed based on some theoretical guidance and existing recommendations for designing efficient genetic representations. Our theoretical and empirical investigations reveal that the S/D binary representation is more proper than other existing binary encodings for searching of genetic algorithms (GAs). Moreover, we define a new genotypic distance on the S/D binary space, which is equivalent to the Euclidean distance on the real-valued space during GAs convergence. Based on the new genotypic distance, GAs can reliably and predictably solve problems of bounded complexity and the methods depended on the Euclidean distance for solving different kinds of optimization problems can be directly used on the S/D binary space.

The toxicity of an anticancer drug is cleared from the body by different processes, including saturable metabolic and nonsaturable renal-excretion pathways. According to the principles of toxicokinetics, we propose a new anticancer drug scheduling model with different toxic elimination processes in this paper. We also present a sophisticated automating drug scheduling approach based on evolutionary computation and computer modeling. To explore multiple efficient drug scheduling policies, we use a multimodal optimization algorithm --- adaptive elitist-population based genetic algorithm (AEGA) to solve the new model, and discuss the situation of multiple optimal solutions under different parameter settings. The simulation results obtained by the new model match well with the clinical treatment experience, and can provide much more drug scheduling policies for a doctor to choose depending on the particular conditions of the patients.

Most genetic programming paradigms are population-based and require huge amount of memory. In this paper, we review the instruction matrix based genetic programming which maintains all program components in a instruction matrix (IM) instead of manipulating a population of programs. A genetic program is extracted from the matrix just before it is being evaluated. After each evaluation, the fitness of the genetic program is propagated to its corresponding cells in the matrix. Then, we extend the instruction matrix to the condition matrix (CM) for generating rule base from datasets. CM keeps some of characteristics of IM and incorporates the information about rule learning. In the evolving process, we adopt an elitist idea to keep the better rules alive to the end. We consider that genetic selection maybe lead to the huge size of rule set, so the reduct theory borrowed from rough sets is used to cut the volume of rules and keep the same fitness as the original rule set. In experiments, we compare the performance of condition matrix for rule learning (CMRL) with other traditional algorithms. Results are presented in detail and the competitive advantage and drawbacks of CMRL are discussed

Kernel-based regularized learning seeks a model in a hypothesis space by minimizing the empirical error and the model's complexity. Based on the representer theorem, the solution consists of a linear combination of translates of a kernel. This paper investigates a generalized form of representer theorem for kernel-based learning. After mapping predefined features and translates of a kernel simultaneously onto a hypothesis space by a specific way of constructing kernels, we proposed a new algorithm by utilizing a generalized regularizer which leaves part of the space unregularized. Using a squared-loss function in calculating the empirical error, a simple convex solution is obtained which combines predefined features with translates of the kernel. Empirical evaluations have confirmed the effectiveness of the algorithm for supervised learning tasks.

This paper presents a novel Genetic Parallel Programming (GPP) paradigm for evolving parallel programs running on a Multi-Arithmetic-Logic-Unit (Multi-ALU) Processor (MAP). The MAP is a Multiple Instruction-streams, Multiple Data-streams (MIMD), general-purpose register machine that can be implemented on modern Very Large-Scale Integrated Circuits (VLSIs) in order to evaluate genetic programs at high speed. For human programmers, writing parallel programs is more difficult than writing sequential programs. However, experimental results show that GPP evolves parallel programs with less computational effort than that of their sequential counterparts. It creates a new approach to evolving a feasible problem solution in parallel program form and then serializes it into a sequential program if required. The effectiveness and efficiency of GPP are investigated using a suite of 14 well-studied benchmark problems. Experimental results show that GPP speeds up evolution substantially.

Various types of integrals with respect to signed fuzzy measures on finite sets with cardinality n can be presented as corresponding rules for partitioning the integrand. The partition can be expressed as an n-dimensional vector, whereas the signed fuzzy measure is also an n-dimensional vector. Thus, the integration value is the inner product of these two vectors. Two pairs of extremes, the Lebesgue-like integral versus the Choquet integral and the upper integral versus the lower integral, are discussed in detail. © 2006 Wiley Periodicals, Inc. Int J Int Syst 21: 1073–1092, 2006.

This paper presents an intelligent virtual environment for Chinese acupuncture learning and training using state-of-the-art virtual reality technology. It is the first step toward developing a comprehensive virtual human model for studying Chinese medicine. Students can learn and practice acupuncture in the proposed 3-D interactive virtual environment that supports a force feedback interface for needle insertion. Thus, students not only "see" but also "touch" the virtual patient. With high performance computers, highly informative and flexible visualization of acupuncture points of various related meridian and collateral can be highlighted to guide the students during training. A computer-based expert system using our newly proposed intelligent fuzzy petri net is designed and implemented to train the students to treat different diseases using acupuncture. Such an intelligent virtual reality system can provide an interesting and effective learning environment for Chinese acupuncture.

In this paper, we introduce a modified optimal control model of drug scheduling in cancer chemotherapy and a new adaptive elitist-population-based genetic algorithm (AEGA) to solve it. Working closely with an oncologist, we first modify the existing model, because its equation for the cumulative drug toxicity is inconsistent with medical knowledge and clinical experience. To explore multiple efficient drug scheduling policies, we propose a novel variable representation--a cycle-wise representation, and modify the elitist genetic search operators in the AEGA. The simulation results obtained by the modified model match well with the clinical treatment experiences, and can provide multiple efficient solutions for oncologists to consider. Moreover, it has been shown that the evolutionary drug scheduling approach is simple, and capable of solving complex cancer chemotherapy problems by adapting multimodal versions of evolutionary algorithms.

This paper proposes a new memetic algorithm (MA) to solve the multi-drug chemotherapy optimization problem. The new MA combines GA with a local search algorithm called iterative dynamic programming (IDP). A multi-drug chemotherapy model is introduced to simulate the possible response of the tumor cells under drugs administration. Optimization of the multiple chemotherapeutic agents' administration schedules is based on this tumor model. We formulate the optimization problem as an optimal control problem (OCP) with a set of dynamic equations. The objective is to design efficient schedules which minimize the tumor size under a set of constraints. Our new MA has been shown to be very efficient on solving our multi-drug model.

Bayesian Network (BN) is a powerful network model, which represents a set of variables in the domain and provides the probabilistic relationships among them. But BN can handle discrete values only; it cannot handle continuous, interval and ordinal ones, which must be converted to discrete values and the order information is lost. Thus, BN tends to have higher network complexity and lower understandability. In this paper, we present a novel dependency network which can handle discrete, continuous, interval and ordinal values through functions; it has lower network complexity and stronger expressive power; it can represent any kind of relationships; and it can incorporate a-priori knowledge though user-defined functions. We also propose a novel Genetic Programming (GP) to learn dependency networks. The novel GP does not use any knowledge-guided nor application-oriented operator, thus it is robust and easy to replicate. The experimental results demonstrate that the novel GP can successfully discover the target novel dependency networks, which have the highest accuracy and the lowest network complexity.

The scalability problem in data mining involves the development of methods for handling large databases with limited computational resources such as memory and computation time. In this paper, two scalable clustering algorithms, bEMADS and gEMADS, are presented based on the Gaussian mixture model. Both summarize data into subclusters and then generate Gaussian mixtures from their data summaries. Their core algorithm, EMADS, is defined on data summaries and approximates the aggregate behavior of each subcluster of data under the Gaussian mixture model. EMADS is provably convergent. Experimental results substantiate that both algorithms can run several orders of magnitude faster than expectation-maximization with little loss of accuracy.

Given the explosive growth of data collected from current business environment, data mining can potentially discover new knowledge to improve managerial decision making. This paper proposes a novel data mining approach that employs an evolutionary algorithm to discover knowledge represented in Bayesian networks. The approach is applied successfully to handle the business problem of finding response models from direct marketing data. Learning Bayesian networks from data is a difficult problem. There are two different approaches to the network learning problem. The first one uses dependency analysis, while the second one searches good network structures according to a metric. Unfortunately, both approaches have their own drawbacks. Thus, we propose a novel hybrid algorithm of the two approaches, which consists of two phases, namely, the conditional independence (CI) test and the search phases. In the CI test phase, dependency analysis is conducted to reduce the size of the search space. In the search phase, good Bayesian network models are generated by using an evolutionary algorithm. A new operator is introduced to further enhance the search effectiveness and efficiency. In a number of experiments and comparisons, the hybrid algorithm outperforms MDLEP, our previous algorithm which uses evolutionary programming (EP) for network learning, and other network learning algorithms. We then apply the approach to two data sets of direct marketing and compare the performance of the evolved Bayesian networks obtained by the new algorithm with those by MDLEP, the logistic regression models, the na/spl inodot//spl uml/ve Bayesian classifiers, and the tree-augmented na/spl inodot//spl uml/ve Bayesian network classifiers (TAN). In the comparison, the new algorithm outperforms the others.

A novel linear genetic programming (LGP) paradigm called genetic parallel programming (GPP) has been proposed to evolve parallel programs based on a multi-ALU processor. It is found that GPP can evolve parallel programs for data classification problems. In this paper, five binary-class UCI machine learning repository databases are used to test the effectiveness of the proposed GPP-classifier. The main advantages of employing GPP for data classification are: 1) speeding up evolutionary process by parallel hardware fitness evaluation; and 2) discovering parallel algorithms automatically. Experimental results show that the GPP-classifier evolves simple classification programs with good generalization performance. The accuracies of these evolved classifiers are comparable to other existing classification algorithms.

We investigate the sample weighting effect on genetic parallel programming (GPP). GPP evolves parallel programs to solve the training samples in a training set. Usually, the samples are captured directly from a real-world system. The distribution of samples in a training set can be extremely biased. Standard GPP assigns equal weights to all samples. It slows down evolution because crowded regions of samples dominate the fitness evaluation causing premature convergence. We present 4 sample weighting (SW) methods, i.e. equal SW, class-equal SW, static SW (SSW) and dynamic SW (DSW). We evaluate the 4 methods on 7 training sets (3 Boolean functions and 4 UCI medical data classification databases). Experimental results show that DSW is superior in performance on all tested problems. In the 5-input symmetry Boolean function experiment, SSW and DSW boost the evolutionary performance by 465 and 745 times respectively. Due to the simplicity and effectiveness of SSW and DSW, they can also be applied to different population-based evolutionary algorithms.

This paper proposes a new approach to combine the knowledge-based model and the cooperation technique of evolutionary agents to identify the location of the desired object in a satellite image. The agents interact with the local information of the image pixels to search for the target objects through an evolutionary process. A new set of fitness function and evolutionary operators are defined for the process. The decentralized, bottom-up and evolutionary natures of the agents can be used to construct a robust system for object recognition in satellite images. The experimental results are satisfactory and have demonstrated the flexibility and power of the approach.

The scalability problem in data mining involves the development of methods for handling large databases with limited computational resources. We present a two-phase scalable model-based clustering framework: first, a large data set is summed up into subclusters; Then, clusters are directly generated from the summary statistics of subclusters by a specifically designed expectation-maximization (EM) algorithm. Taking example for Gaussian mixture models, we establish a provably convergent EM algorithm, EMADS, which embodies cardinality, mean, and covariance information of each subcluster explicitly. Combining with different data summarization procedures, EMADS is used to construct two clustering systems: gEMADS and bEMADS. The experimental results demonstrate that they run several orders of magnitude faster than the classic EM algorithm with little loss of accuracy. They generate significantly better results than other model-based clustering systems using similar computational resources.

Abstract This paper presents a stochastic dynamic load balancing algorithm for Internet computing, which is a new type of distributed computing involving heterogeneous workstations from different organizations on the Internet. To realize the practical environment, we assume the system to be comprised of heterogeneous, untrusted and non‐dedicated workstations connected by a non‐dedicated network. Our algorithm uses the product of the average processing time and the queue length of system jobs as the load index. Dynamic communication delay is included in the execution cost calculation. The transfer policy and the location policy are combined in a stochastic algorithm. State information exchange is done via information feedback and mutual updating. Simulations demonstrate that our algorithm outperforms conventional approaches over a wide range of system parameters. These results are reconfirmed by empirical experiments after we have implemented the algorithms on the Distributed Java Machine global virtual machine. Copyright © 2003 John Wiley & Sons, Ltd.

Abstract Many real‐world optimization problems in the scientific and engineering fields can be solved by genetic algorithms (GAs) but it still requires a long execution time for complex problems. At the same time, there are many under‐utilized workstations on the Internet. In this paper, we present a self‐adaptive parallel GA system named APGAIN, which utilizes the spare power of the heterogeneous workstations on the Internet to solve complex optimization problems. In order to maintain a balance between exploitation and exploration, we have devised a novel probabilistic rule‐driven adaptive model (PRDAM) to adapt the GA parameters automatically. APGAIN is implemented on an Internet Computing system called DJM. In the implementation, we discover that DJM's original load balancing strategy is insufficient. Hence the strategy is extended with the job migration capability. The performance of the system is evaluated by solving the traveling salesman problem with data from a public database. Copyright © 2003 John Wiley & Sons, Ltd.

A new method based on nonlinear integral projections for classification is presented. The contribution rate of each combination of the feature attributes, including each singleton, toward the classification is represented by a fuzzy measure. The nonadditivity of the fuzzy measure reflects the interactions among the feature attributes. The weighted Choquet integral with respect to the fuzzy measure serves as an aggregation tool to project the feature space onto a real axis optimally according to an error criterion, and the classifying attribute is properly numerical analysed on the axis simultaneously making the classification simple. To implement the classification, we need to determine the unknown parameters, the values of fuzzy measure and the weight function. This can be done by running an adaptive genetic algorithm on the given training data. The new classifier is tested by recovering the preset parameters from a set of artificial training data generated from these parameters. It also performs well on several real-world data sets. Beyond discriminating classes, this method can also learn the scaling requirements and the respective importance indexes of the feature attributes as well as the relationships among them. A comprehensive discussion on the semantic and geometric meanings of the parameters is given. Moreover, we show how these parameters' values can be used for short-listing important feature attributes to reduce the complexity (dimensions) of the classification problem. Our new method also compares favorably with other methods on some well-known real-world benchmarks.

As a typical combinatorial optimization problem, the traveling salesman problem (TSP) has attracted extensive research interest. In this paper, we develop a self-organizing map (SOM) with a novel learning rule. It is called the integrated SOM (ISOM) since its learning rule integrates the three learning mechanisms in the SOM literature. Within a single learning step, the excited neuron is first dragged toward the input city, then pushed to the convex hull of the TSP, and finally drawn toward the middle point of its two neighboring neurons. A genetic algorithm is successfully specified to determine the elaborate coordination among the three learning mechanisms as well as the suitable parameter setting. The evolved ISOM (eISOM) is examined on three sets of TSP to demonstrate its power and efficiency. The computation complexity of the eISOM is quadratic, which is comparable to other SOM-like neural networks. Moreover, the eISOM can generate more accurate solutions than several typical approaches for TSP including the SOM developed by Budinich, the expanding SOM, the convex elastic net, and the FLEXMAP algorithm. Though its solution accuracy is not yet comparable to some sophisticated heuristics, the eISOM is one of the most accurate neural networks for the TSP.

This paper proposes a novel genetic parallel programming (GPP) paradigm for evolving optimal parallel programs running on a multi-ALU processor by linear genetic programming. GPP uses a two-phase evolution approach. It evolves completely correct solution programs in the first phase. Then it optimizes execution speeds of solution programs in the second phase. Besides, GPP also employs a new genetic operation that swaps sub-instructions of a solution program. Three experiments (Sextic, Fibonacci and Factorial) are given as examples to show that GPP could discover novel parallel programs that fully utilize the processor's parallelism.

In this paper, a new two-way adaptive search mutation strategy is proposed, and a "two-way evolution strategy" (TWES) is established. The experimental results show that TWES yields much faster convergence than classical evolution strategies. This paper also discusses the relationship between the parameter setting and the convergent speed by TWES.

A novel hybrid framework is reported that improves upon our previous work, MDLEP, which uses evolutionary programming to solve the difficult Bayesian network learning problem. A new merge operator is also introduced that further enhances the efficiency. As experimental results suggest, our hybrid approach performs significantly better than MDLEP.

In this paper, we present a new algorithm-asynchronous self-adjustable island genetic algorithm (aSAIGA) for multi-objective optimization problems. The proposed algorithm is built upon the coarse-grained architecture, which is divided into sub-processes and distributed amongst several island processors. In each sub-process, an asynchronous communication operation and a self-adjusting operation are adopted to enhance the algorithm in both speedup and global searching capabilities. Satisfactory results and significant speedup can be achieved by aSAIGA, as shown by simulation.

In this paper, we propose a complete and efficient algorithm to perform the fuzzy inference of a rule-based system modeled by fuzzy Petri nets (FPN). The algorithm is developed to simulate the inference process from the starting propositions to the goal propositions. The structures of the FPN are in more generic forms, compared to those applied in previous papers. The formal description of models and the fuzzy reasoning algorithm are described in detail, with several illustration examples. The differences between other algorithms are also presented.

The self-organizing map (SOM) is a powerful tool in the exploratory phase of data mining. However, due to the dimensional conflict, neighborhood preservation cannot always lead to perfect topology preservation. In this paper we establish an expanding SOM (ESOM) to detect and preserve better topology correspondence between the two spaces. Our experiment results demonstrate that the ESOM constructs better mappings than the classic SOM in terms of both topological and quantization errors. Furthermore, clustering results generated by the ESOM are more accurate than those of the SOM.

Describes a data mining approach that employs evolutionary programming to discover knowledge represented in Bayesian networks. There are two different approaches to the network learning problem. The first one uses dependency analysis, while the second one searches good network structures according to a metric. Unfortunately, both approaches have their own drawbacks. Thus, we propose a hybrid algorithm of the two approaches, which consists of two phases, namely, the conditional independence test and the search phases. A new operator is introduced to further enhance the search efficiency. We conduct a number of experiments and compare the hybrid algorithm with our previous algorithm, MDLEP, which uses EP for network learning. The empirical results illustrate that the new approach has better performance. We apply the approach to data sets of direct marketing and compare the performance of the evolved Bayesian networks obtained by the new algorithm with the models generated by other methods. In the comparison, the induced Bayesian networks produced by the new algorithm outperform the other models.

This paper addresses a bottleneck problem in mobile file systems: the propagation of updated large tiles from a weakly-connected client to its servers. It proposes an efficient mechanism called operation shipping or operation-based update propagation. In the new mechanism, the client ships the user operation that updated the large files, rather than the files themselves, across the weak network. (in contrast, existing file systems use value shipping and ship the files.) The user operation is sent to a surrogate client that is strongly connected to the servers. The surrogate replays the user operation, regenerates the files, checks whether they are identical to the originals, and, it so, sends the files to the servers on behalf of the client. Care has been taken such that the new mechanism does not compromise correctness or server scalability. For example, we show how forward error correction (FEC) can restore minor reexecution discrepancies and, thus, make operation shipping work with more applications. Operation shipping can be further classified into two types: application-transparent and application-aware. Their feasibilities and benefits have been demonstrated by the design, implementation, and evaluation of a prototype extension to the Coda File System. In our controlled experiments, operation shipping achieved substantial performance improvements-network traffic reductions from 12 times to nearly 400 times and speedups in the range of 1.4 times to nearly 50 times.

This paper describes a novel knowledge discovery and data mining framework dealing with nonlinear interactions among domain attributes. Our network-based model provides an effective and efficient reasoning procedure to perform prediction and decision making. Unlike many existing paradigms based on linear models, the attribute relationship in our framework is represented by nonlinear nonnegative multiregressions based on the Choquet integral. This kind of multiregression is able to model a rich set of nonlinear interactions directly. Our framework involves two layers. The outer layer is a network structure consisting of network elements as its components, while the inner layer is concerned with a particular network element modeled by Choquet integrals. We develop a fast double optimization algorithm (FDOA) for learning the multiregression coefficients of a single network element. Using this local learning component and multiregression-residual-cost evolutionary programming (MRCEP), we propose a global learning algorithm, called MRCEP-FDOA, for discovering the network structures and their elements from databases. We have conducted a series of experiments to assess the effectiveness of our algorithm and investigate the performance under different parameter combinations, as well as sizes of the training data sets. The empirical results demonstrate that our framework can successfully discover the target network structure and the regression coefficients.

Multiregression is one of the most common approaches used to discover dependency pattern among attributes in a database. Nonadditive set functions have been applied to deal with the interactive predictive attributes involved, and some nonlinear integrals with respect to nonadditive set functions are employed to establish a nonlinear multiregression model describing the relation between the objective attribute and predictive attributes. The values of the nonadditive set function play a role of unknown regression coefficients in the model and are determined by an adaptive genetic algorithm from the data of predictive and objective attributes. Furthermore, such a model is now improved by a new numericalization technique such that the model can accommodate both categorical and continuous numerical attributes. The traditional dummy binary method dealing with the mixed type data can be regarded as a very special case of our model when there is no interaction among the predictive attributes and the Choquet integral is used. When running the algorithm, to avoid a premature during the evolutionary procedure, a technique of maintaining diversity in the population is adopted. A test example shows that the algorithm and the relevant program have a good reversibility for the data. © 2001 John Wiley & Sons, Inc.16: 949–962 (2001)

There have been various algorithms designed for simulating natural evolution. This paper proposes a new simulated evolutionary computation model called the abstract evolutionary algorithm (AEA), which unifies most of the currently known evolutionary algorithms and describes the evolution as an abstract stochastic process composed of two fundamental operators: selection and evolution operators. By axiomatically characterizing the properties of the fundamental selection and evolution operators, several general convergence theorems and convergence rate estimations for the AEA are established. The established theorems are applied to a series of known evolutionary algorithms, directly fielding new convergence conditions and convergence rate estimations of various specific genetic algorithms and evolutionary strategies. The present work provides a significant step toward the establishment of a unified theory of simulated evolutionary computation.

A new gradient-based neural network is constructed on the basis of the duality theory, optimization theory, convex analysis theory, Lyapunov stability theory, and LaSalle invariance principle to solve linear and quadratic programming problems. In particular, a new function F(x, y) is introduced into the energy function E(x, y) such that the function E(x, y) is convex and differentiable, and the resulting network is more efficient. This network involves all the relevant necessary and sufficient optimality conditions for convex quadratic programming problems. For linear programming and quadratic programming (QP) problems with unique and infinite number of solutions, we have proven strictly that for any initial point, every trajectory of the neural network converges to an optimal solution of the QP and its dual problem. The proposed network is different from the existing networks which use the penalty method or Lagrange method, and the inequality constraints are properly handled. The simulation results show that the proposed neural network is feasible and efficient.

In data mining, we emphasize the need for learning from huge, incomplete, and imperfect data sets. To handle noise in the problem domain, existing learning systems avoid overfitting the imperfect training examples by excluding insignificant patterns. The problem is that these systems use a limiting attribute-value language for representing the training examples and the induced knowledge. Moreover, some important patterns are ignored because they are statistically insignificant. In this article, we present a framework that combines Genetic Programming and Inductive Logic Programming to induce knowledge represented in various knowledge representation formalisms from noisy databases. The framework is based on a formalism of logic grammars, and it can specify the search space declaratively. An implementation of the framework, LOGENPRO (The Logic grammar based GENetic PROgramming system), has been developed. The performance of LOGENPRO is evaluated on the chess end-game domain. We compare LOGENPRO with FOIL and other learning systems in detail, and find its performance is significantly better than that of the others. This result indicates that the Darwinian principle of natural selection is a plausible noise handling method that can avoid overfitting and identify important patterns at the same time. Moreover, the system is applied to one real-life medical database. The knowledge discovered provides insights to and allows better understanding of the medical domains.

Simulated annealing is a robust and easy-to-implement algorithm for material simulation. However, it consumes a huge amount of computational time, especially on the studies of percolation networks. To reduce the running time, we parallelize the simulated annealing algorithm in our studies of the thermoelastic scaling behavior of percolation networks. The critical properties of the thermoelastic moduli of percolation networks near the threshold p/sub c/ are investigated by constructing a square percolation network. The properties are tested by simulations of a series of two-dimensional (2-D) percolation networks near p/sub c/. The simulations are performed using a novel parallelizing scheme on the simulated annealing algorithm. To further accelerate the computational speed, we also propose a new conjectural method to generate better initial configurations, which speeds up the simulation significantly. Preliminary simulation results show surprisingly that the percolating phenomenon of thermal expansion does exist under certain conditions. The behavior seems to be governed by the elastic properties of a percolation network.

Unlike most of the current object-oriented geographical information systems (OOGISs) whose designs are based on the traditional spatial conceptual model emphasizing the processing of geometric features, the concept-based OOGIS proposed in this paper provides a spatial conceptual model which comprises rich spatial semantics fundamental to spatial analysis, and an object-oriented data model (OODM) which provides an appropriate and effective representation of the spatial conceptual model for efficient database management. By structuring the cognition of space through three interrelated hierarchies: namely the spatial conceptual hierarchy, the entity hierarchy and the feature hierarchy, the generic concept-based OOGIS renders an appropriate framework for the scientific investigation of space and the design of an efficient object-oriented database management system. In addition to its generic nature, the proposed OOGIS is in line withour commonsense conceptualization of space. Furthermore, it can entertain multiple-representations, and can facilitate data integration and generalization. The present investigation thus advances an effective way for OOGIS research in general and design in particular.

We have developed a new approach to learning Bayesian network structures based on the minimum description length (MDL) principle and evolutionary programming. It employs a MDL metric, which is founded on information theory, and integrates a knowledge-guided genetic operator for the optimization in the search process.

We propose a new load distribution (LD) algorithm which is based on anti-tasks and load state vectors. Anti-tasks are composite agents which travel around a distributed system to facilitate the pairing up of task senders and receivers, as well as the collection and dissemination of load information. Time-stamped load information of processing nodes is stored in load state vectors which, when used together with anti-tasks, encourage mutual sharing of load information among processing nodes. Anti-tasks, which make use of load state vectors to decide their travelling paths, are spontaneously directed towards processing nodes having high transient workload, thus allowing their surplus work-load to be relocated quickly.

Existing search-based discrete global optimization methods share two characteristics: 1) searching at the highest resolution; and 2) searching without memorizing past searching information. In this paper, we provide a model which copes: 1) structurally, it transforms the optimization problem into a selection problem by organizing the continuous search space into a binary hierarchy of partitions; and 2) algorithmically, it is an iterative stochastic cooperative-competitive searching algorithm with memory. It is pointed out that the competition model eliminates the requirement of the niche radius required in the existing niching techniques. The model is applied to (but not limited to) function optimization problems (including high-dimensional problems) with experimental results which show that our model is promising for global optimization. We show how pccBHS can be integrated into genetic algorithms as an operator.

By using a non-additive set function to describe the interaction among variables, a nonlinear non-negative multi-regression is established based on the Choquet integral with respect to the set function. We generalize this nonlinear model and propose a novel formalism that provides an effective and efficient reasoning procedure to perform information fusion, decision making, and medical diagnoses. In the formalism, a network structure and a number of Choquet integrals are used to represent the relationships among variables. We propose a new algorithm to learn the network structure and the regression parameters of Choquet integrals from training examples in databases. The algorithm is based on the minimum description length (MDL) principle and evolutionary programming (EP). We conduct a series of experiments to demonstrate the performance of our algorithm and estimate the effectiveness of the MDL metric and the genetic operators. The empirical results illustrate that our algorithm can successfully discover the target network structure and the regression parameter.

We develop a nonlinear multi-regression model based on the Wang integral to describe a multi-input single-output system. In this model, in general, set function /spl mu/ is nonadditive. The nonadditivity of /spl mu/ describes the inherent interaction among the input attributes x/sub 1/, x/sub 2/, ..., x/sub n/. When the proper input-output data are available, by using the adaptive genetic algorithm shown in this paper, rather precise estimated values of parameter c, q, w and /spl mu/ of the regression model can be obtained. Thus, the multi-input single-output system can be used to make prediction. That is to say, when the values of input attributes x/sub 1/, X/sub 2/, ..., X/sub n/, are known, we can predict the output Y by calculating the nonlinear multi-regression.

Polling-based load distribution (LD) algorithms suffer from two weaknesses: (i) load information exchanged during a polling session is confined to the two negotiating nodes only; (ii) as the distributed system grows in size (in terms of the number of constituent nodes), a larger number of polling sessions, and thus a higher amount of network bandwidth consumption and CPU overhead, are needed. We propose a new LD algorithm which is based on anti-tasks and load state vectors. This new algorithm avoids the above weaknesses of polling-based LD algorithms. Anti-tasks are composite agents which travel around a distributed system to facilitate the pairing up of task senders and receivers, as well as the collection and dissemination of load information. Time-stamped load information of processing nodes is stored in load state vectors which, when used together with anti-tasks, encourage mutual sharing of load information among processing nodes. Anti-tasks, which make use of load state vectors to decide their traveling paths, are spontaneously directed towards processing nodes having high transient workload, thus allowing their surplus workload to be relocated quickly. Using simulations, we evaluate the performance of our new algorithm by comparing its performance with a number of well-known polling-based load distribution algorithms. We found that our algorithm provides significant reduction of mean task response time over a large range of system sizes. The cost of achieving this performance gain in terms of CPU overhead and channel bandwidth consumption is generally comparable to the other algorithms we studied. © 1998 John Wiley & Sons, Ltd.

In this paper, we present the design and implementation of a novel model of distributed computing on the Internet and Intranet environment. Our model is called the Distributed Java Machine (DJM). It is a global distributed virtual machine used to realize the concept of ‘network is the computer’. DJM explores coarse-grained parallelism by using the under-utilized workstations on the network, combining the elements of object-oriented technology, distributed computing, World-Wide Web and Java programming. It can run on machines with heterogeneous hardware and software platforms without relinking or recompilation. DJM has two unique features. First, using an original applet helper mechanism, DJM allows machines without any DJM software and of different levels of ‘trust’ to work together. Secondly, DJM has implemented concurrency enhancement mechanisms (one-way message, future and redirected future) to increase the efficiency of method invocation. The prototype of DJM has been implemented and tested under both the Intranet and the Internet environments. Using the workstations from our teaching laboratories, which are already running under normal loading, experimental results show that we can achieve a speedup of about 5–8 times by 14 workstations in a Local Area Network (LAN) environment, and about 4.5 times speedup for eight workstations in a Wide Area Network (WAN) environment. © 1998 John Wiley & Sons, Ltd.

Program induction generates a computer program that can produce the desired behavior for a given set of situations. Two of the approaches in program induction are inductive logic programming (ILP) and genetic programming (GP). Since their formalisms are so different, these two approaches cannot be integrated easily, although they share many common goals and functionalities. A unification will greatly enhance their problem-solving power. Moreover, they are restricted in the computer languages in which programs can be induced. In this paper, we present a flexible system called LOGENPRO (The LOgic gramar-based GENetic PROgramming system) that uses some of the techniques of GP and ILP. It is based on a formalism of logic grammars. The system applies logic grammars to control the evolution of programs in various programming languages and represent context-sensitive information and domain-dependent knowledge. Experiments have been performed to demonstrate that LOGENPRO can emulate GP and GP with automatically defined functions (ADFs). Moreover, LOGENPRO can employ knowledge such as argument types in a unified framework. The experiments show that LOGENPRO has superior performance to that of GP and GP with ADFs when more domain-dependent knowledge is available. We have applied LOGENPRO to evolve general recursive functions for the even-n-parity problem from noisy training examples. A number of experiments have been performed to determine the impact of domain-specific knowledge and noise in training examples on the speed of learning.

The authors consider optimal placements of two-dimensional flexible (elastic, deformable) objects. The objects are discs of equal size placed within a rigid boundary. The paper is divided into two parts. In the first part, analytical results for three types of regular, periodic arrangements-the hexagonal, square, and triangular placements-are presented. The regular arrangements are analyzed for rectangular boundaries and radii of discs that are small compared to the area of the placement region, because, in this case, the influence of boundary conditions can be neglected. This situation is called the unbounded case. They show that, for the unbounded case among the three regular placements, the type of hexagonal arrangements provides the largest number of placed units for the same deformation depth. Furthermore, it can be proved that these regular placements are not too far from the truly optimal arrangements. For example, hexagonal placements differ at most by the factor of 1.1 from the largest possible number of generally shaped units in arbitrary arrangements. These analytical results are used as guidances for testing stochastic algorithms optimizing placements of flexible objects. In the second part, mainly two problems are considered: the underlying physical model and a simulated annealing algorithm maximizing the number of flexible discs in equilibrium placements. Along with the physical model, an approximate formula is derived, reflecting the deformation/force relationship for a large range of deformations.

For pt.I see ibid., p.890-904. The paper is a continuation of the first part, where the authors considered regular arrangements of flexible objects for the unbounded case. The present part deals with a simulated annealing algorithm maximizing the number of flexible objects in equilibrium placements within rigid boundaries. The forces caused by the boundary are taken into account, i.e., the bounded case of placements is considered. The simulated annealing procedure makes use of the special structure of the underlying configuration space and relationships between deformations of flexible objects and resulting forces. This allows one to obtain tight bounds for the annealing parameters which result in n/sup 3/2//spl middot/In/sup 5/2/ and n/spl middot/In/sup 2/n time bounds, respectively, for the computation of equilibrium states by two different cooling schedules. The deformation/force formula is derived from a physical model of flexible discs and is based on numerical experiments which were performed for different materials and different sizes of objects. The algorithm was first implemented and tested for the unbounded case. The run-time is relatively short, even for large numbers of placed discs. These results are compared to the analytical ones obtained for regular placements in the first part of the paper, and agreement between these two sets of results are observed. Furthermore, several experiments for placements with boundary conditions were carried out and the resulting placements clearly show the effect of the forces from the rigid boundary.

Associative-memory neural networks with adaptive weighted outer-product learning are proposed in this paper. For the correct recall of a fundamental memory (FM), a corresponding learning weight is attached and a parameter called signal-to-noise-ratio-gain (SNRG) is devised. The sufficient conditions for the learning weights and the SNRG's are derived. It is found both empirically and theoretically that the SNRG's have their own threshold values for correct recalls of the corresponding FM's. Based on the gradient-descent approach, several algorithms are constructed to adaptively find the optimal learning weights with reference to global- or local-error measure.

Genetic programming (GP) and inductive logic programming (ILP) have received increasing interest. Since their formalisms are so different these two approaches cannot be integrated easily though they share many common goals and functionalities. A unification will greatly enhance their problem solving power. Moreover, they are restricted in the computer languages in which programs can be induced. We present a flexible system called LOGENPRO (The logic grammar based genetic programming system) that combines GP and ILP. It is based on a formalism of logic grammars. The system can learn programs in various programming languages and represent context-sensitive information and domain-dependent knowledge. The performance of LOGENPRO in inducing logic programs from noisy examples is evaluated. A detailed comparison with FOIL has been conducted. This experiment demonstrates that LOGENPRO is a promising alternative to other inductive logic programming systems and sometimes is superior for handling noisy data. Moreover, a series of examples are used to illustrate that LOGENPRO is so flexible that programs in different programming languages including LISP, Prolog and Fuzzy Prolog can be induced.

An intelligent expert system shell for the development of knowledge-based geographical information systems (GIS) is examined in this two-part article. Basic concepts and the overall architecture of the shell are discussed in the present part. Fuzzy logic and expert systems technology are demonstrated to be appropriate methods for approximating human reasoning and enhancing the level of intelligence in GIS. The shell can be employed as an effective and efficient tool for developing knowledge-based GIS.

Employing the expert system shell discussed in part 1 of this two–part article, three simple geographical expert systems are constructed as didactic examples. The first deals with land–type classification with remotely–sensed data, the second with climatic classifications with regular data files, and the third involves discussions on building DTM–related expert systems. All are rule–based expert systems easily built from the shell. It is apparent that the shell provides a powerful software environment for developing knowledge–based GIS. Directions for further research are also outlined.

Fuzzy logic is one of the methods to model the vagueness and imprecision of human knowledge. Some rule-based expert system shells have been successfully developed and have demonstrated the power of fuzzy logic in dealing with inexact reasoning and rule inferences. However, using rules for knowledge representation is not structured enough. In addition, knowledge cannot be easily represented in an abstracted (hierarchical) from. In this article the introduction of fuzzy concepts into object oriented knowledge representation (OOKR), which is a structured knowledge representation scheme, is presented. A framework for handling all the possible fuzzy concepts in OOKR at both the dynamic and static levels is proposed. In order to handle the inheritance mechanism and to model the relations among classes, instances, and attributes, some new fuzzy concepts and operations are introduced. These concepts and operations are developed from the semantic meaning rather than by an ad hoc approach. A prototype of the expert system shell. System FX-I, has been successfully developed based on the above framework, showing the feasibility of handling inexact knowledge in a structural way.