This paper proposes a NPEL (Nine-Pipeline & Ensemble Learning) strategy based on machine Learning algorithm to predict protein-protein interaction hotspots by training amino acid composition, surface area, amino acid chains and other complex/interface-related structural information. We applied Random Forest, Linear Svm, KNN, Gaussian Naive Bayes, Multi-layer Perceptron Neural Network, Adaboost, XGBoost etc. nine machine learning algorithms combination into an independent pipeline to predict protein hot spots, and the final results are optimized through voting and stacking scheme. In the stacking result of XGBoost and Logistic Regression, the highest accuracy is 0.8462 and improve the indicators of the pipeline results greatly.

BACKGROUND: In the process of designing drugs and proteins, it is crucial to recognize hot regions in protein-protein interactions. Each hot region of protein-protein interaction is composed of at least three hot spots, which play an important role in binding. However, it takes time and labor force to identify hot spots through biological experiments. If predictive models based on machine learning methods can be trained, the drug design process can be effectively accelerated. RESULTS: The results show that different machine learning algorithms perform similarly, as evaluating using the F-measure. The main differences between these methods are recall and precision. Since the key attribute of hot regions is that they are packed tightly, we used the cluster algorithm to predict hot regions. By combining Gaussian Naïve Bayes and DBSCAN, the F-measure of hot region prediction can reach 0.809. CONCLUSIONS: In this paper, different machine learning models such as Gaussian Naïve Bayes, SVM, Xgboost, Random Forest, and Artificial Neural Network are used to predict hot spots. The experiment results show that the combination of hot spot classification algorithm with higher recall rate and clustering algorithm with higher precision can effectively improve the accuracy of hot region prediction.

Discovering hot regions in protein-protein interaction is important for understanding the interactions between proteins, while because of the complexity and time-consuming of experimental methods, the computational prediction method can be very helpful to improve the efficiency to predict hot regions. In previous researches, some models are based on a single aspect, such as structure, energy, and sequence, each aspect has its advantage and limitations. In this paper, a new method that combing structure-based classification, energy-based clustering and sequence-based conservation in evolution is proposed. This method makes full use of three aspects of protein information and compensates for the limitations of using one single aspect information. Experimental results show that the proposed method significantly improves the prediction performance of hot regions.

SUMMARY: FeatureStack is a Perl module for the automatic generation of multi-gene images. FeatureStack takes BioPerl-compliant gene or transcript features as input and renders them side by side using a user-defined BioPerl glyph. Output images can be generated in SVG or PNG format. FeatureStack comes with a new BioPerl glyph, decorated_gene, which can highlight protein features on top of gene models. Used in combination, FeatureStack and decorated_gene enable rapid and automated generation of annotation-rich images of stacked gene models that greatly facilitate evolutionary studies of related gene structures and gene families. AVAILABILITY AND IMPLEMENTATION: Bio-Draw-FeatureStack and Bio-Graphics-glyph-decorated_gene are freely available at the Comprehensive Perl Archive Network (CPAN) and GitHub. CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: BLAST users frequently expect to obtain homologous genes with certain similarity to their query genes. But what they get from BLAST searches are often collections of local alignments called high-scoring segment pairs (HSPs). On the other hand, most homology-based gene finders have been built using computation-intensive algorithms, without taking full advantage of BLAST searches that have been perfected over the last decades. RESULTS: Here we report an efficient algorithm, genBlastG that directly uses the HSPs reported by BLAST to define high-quality gene models. AVAILABILITY: http://genome.sfu.ca/genblast/download.html

Genes underlying important phenotypic differences between Plasmodium species, the causative agents of malaria, are frequently found in only a subset of species and cluster at dynamically evolving subtelomeric regions of chromosomes. We hypothesized that chromosome-internal regions of Plasmodium genomes harbour additional species subset-specific genes that underlie differences in human pathogenicity, human-to-human transmissibility, and human virulence. We combined sequence similarity searches with synteny block analyses to identify species subset-specific genes in chromosome-internal regions of six published Plasmodium genomes, including Plasmodium falciparum, Plasmodium vivax, Plasmodium knowlesi, Plasmodium yoelii, Plasmodium berghei, and Plasmodium chabaudi. To improve comparative analysis, we first revised incorrectly annotated gene models using homology-based gene finders and examined putative subset-specific genes within syntenic contexts. Confirmed subset-specific genes were then analyzed for their role in biological pathways and examined for molecular functions using publicly available databases. We identified 16 genes that are well conserved in the three primate parasites but not found in rodent parasites, including three key enzymes of the thiamine (vitamin B1) biosynthesis pathway. Thirteen genes were found to be present in both human parasites but absent in the monkey parasite P. knowlesi, including genes specifically upregulated in sporozoites or gametocytes that could be linked to parasite transmission success between humans. Furthermore, we propose 15 chromosome-internal P. falciparum-specific genes as new candidate genes underlying increased human virulence and detected a currently uncharacterized cluster of P. vivax-specific genes on chromosome 6 likely involved in erythrocyte invasion. In conclusion, Plasmodium species harbour many chromosome-internal differences in the form of protein-coding genes, some of which are potentially linked to human disease and thus promising leads for future laboratory research.

Gene prediction is one of the most challenging tasks in genome analysis, for which many tools have been developed and are still evolving. In this paper, we present a novel gene prediction method that is both fast and accurate, by making use of protein homology and decision tree classification. Specifically, we apply the principled entropy and decision tree concepts to assist in such gene prediction process. Our goal is to resolve the exact gene structures in terms of finding “coding” regions (exons) and “non-coding” regions (introns). Unlike traditional classification tasks, however, we do not have explicit class labels for such structures in the genes. We use protein sequence (the product of gene) as a query to help in finding genes that are homologous to the query protein and deduce class labels based on homology. Our experiments on the genomes of two nematodes C. elegans and C. briggsae show that in addition to achieving prediction accuracy comparable with that of the state of the art methods, it is several orders of magnitude faster, especially for genes that encode longer proteins.

BACKGROUND: The recent availability of an expanding collection of genome sequences driven by technological advances has facilitated comparative genomics and in particular the identification of synteny among multiple genomes. However, the development of effective and easy-to-use methods for identifying such conserved gene clusters among multiple genomes-synteny blocks-as well as databases, which host synteny blocks from various groups of species (especially eukaryotes) and also allow users to run synteny-identification programs, lags behind. DESCRIPTIONS: OrthoClusterDB is a new online platform for the identification and visualization of synteny blocks. OrthoClusterDB consists of two key web pages: Run OrthoCluster and View Synteny. The Run OrthoCluster page serves as web front-end to OrthoCluster, a recently developed program for synteny block detection. Run OrthoCluster offers full control over the functionalities of OrthoCluster, such as specifying synteny block size, considering order and strandedness of genes within synteny blocks, including or excluding nested synteny blocks, handling one-to-many orthologous relationships, and comparing multiple genomes. In contrast, the View Synteny page gives access to perfect and imperfect synteny blocks precomputed for a large number of genomes, without the need for users to retrieve and format input data. Additionally, genes are cross-linked with public databases for effective browsing. For both Run OrthoCluster and View Synteny, identified synteny blocks can be browsed at the whole genome, chromosome, and individual gene level. OrthoClusterDB is freely accessible. CONCLUSION: We have developed an online system for the identification and visualization of synteny blocks among multiple genomes. The system is freely available at (http://genome.sfu.ca/orthoclusterdb/).

By comparing genomes among both closely and distally related species, comparative genomics analysis characterizes structures and functions of different genomes in both conserved and divergent regions. Synteny blocks, which are conserved blocks of genes on chromosomes of related species, play important roles in comparative genomics analysis. Although a few tools have been designed to identify synteny blocks, most of them cannot handle some challenging application requirements, particularly the strandedness of genes, gene inversions, gene duplications, and comparison of more than two genomes. We developed a data mining tool, Ortho-Cluster, which can handle all those challenges. It is publicly available at http://genome.sfu.ca/projects/orthocluster. OrthoCluster takes the annotated gene sets of candidate genomes and pairwise orthologous relationships as input and efficiently identifies the complete set of synteny blocks. In addition, OrthoCluster identifies four types of genome rearrangement events namely inversion, transposition, insertion/deletion, and reciprocal translocation. To be fleexible in various application scenarios, OrthoCluster comes with a systematic set of parameters such as the synteny block size, number of mismatches allowed, whether the strandedness is enforced, whether gene ordering is preserved. Furthermore, OrthoCluster can be used to identify segmental duplication in a genome. In this paper, we introduce the major technical ideas, and present some interesting findings using OrthoCluster.