MOTIVATION: Complex gene regulatory networks (GRNs) underlie most disease processes, and understanding disease-specific network structures and dynamics is crucial for developing effective treatments. Yet, most database- and literature-based analyses of GRNs often treat gene regulations as context-independent interactions, overlooking how GRNs can differ depending on the disease type, cell lineage, or experimental condition. RESULTS: In an attempt to improve on existing methods for leveraging knowledge present in the scientific literature, we developed a framework to assign quantitative, context-dependent weights to gene regulations extracted from literature. We demonstrate that the context-specific GRNs reconstructed with our method can effectively capture disease biology, showing strong correlation with transcriptomics across a wide range of diseases. Furthermore, we show that utilizing contextual information improves accuracy in drug-target prediction tasks. Finally, we showcase the utility of the contextualized GRNs through the automated construction of an ordinary differential equation model of a breast cancer-specific signaling network. The large language model-based framework allows the integration of literature- and experimentally derived information and streamlines the process of assembling a biologically relevant and functional mathematical model. Our findings indicate the importance of considering the context when making biological predictions, and we demonstrate the use of natural language processing tools to effectively mine associations between gene regulations and biological contexts. AVAILABILITY AND IMPLEMENTATION: All reproducibility code is available at https://github.com/okadalabipr/context-dependent-GRNs, along with the automated mathematical model construction package at https://github.com/okadalabipr/BioMathForge. The dataset used in this study is available at Zenodo, DOI: 10.5281/zenodo.16416117.

MOTIVATION: Single-cell RNA sequencing (scRNA-seq) analysis reveals heterogeneity and dynamic cell transitions. However, conventional gene-based analyses require intensive manual curation to interpret biological implications of computational results. Hence, a theory for efficiently annotating individual cells remains warranted. RESULTS: We present ASURAT, a computational tool for simultaneously performing unsupervised clustering and functional annotation of disease, cell type, biological process and signaling pathway activity for single-cell transcriptomic data, using a correlation graph decomposition for genes in database-derived functional terms. We validated the usability and clustering performance of ASURAT using scRNA-seq datasets for human peripheral blood mononuclear cells, which required fewer manual curations than existing methods. Moreover, we applied ASURAT to scRNA-seq and spatial transcriptome datasets for human small cell lung cancer and pancreatic ductal adenocarcinoma, respectively, identifying previously overlooked subpopulations and differentially expressed genes. ASURAT is a powerful tool for dissecting cell subpopulations and improving biological interpretability of complex and noisy transcriptomic data. AVAILABILITY AND IMPLEMENTATION: ASURAT is published on Bioconductor (https://doi.org/10.18129/B9.bioc.ASURAT). The codes for analyzing data in this article are available at Github (https://github.com/keita-iida/ASURATBI) and figshare (https://doi.org/10.6084/m9.figshare.19200254.v4). SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Genetic network inference methods using random forests have shown promise. Some of the random-forest-based inference methods have an ability to analyze both time-series and static gene expression data. We think however that, as the gene expression levels at two adjacent measurements of a time-series data are often similar to each other, the gene expression levels at each measurement in the time-series data are less informative than those in the static data. On the basis of this idea, we proposed a new inference method that relies more on static gene expression data than time-series ones. Through the numerical experiments, we showed that the quality of the inferred genetic network is slightly improved by giving greater importance to static data than time-series ones. Although we develop the new method by modifying the random-forest-based inference method proposed by the authors, we could introduce the idea in this study into any inference method that is capable of analyzing both time-series and static gene expression data.

Huynh-Thu and colleagues initially introduce the random forest into field of genetic network inference. Their method, GENIE3, has performed well on genetic network inference problems. However, GENIE3 was designed only for analyzing static expression data that were measured under steady-state conditions. In order to infer genetic networks from time-series of gene expression data, this study proposes a new method based on the random forest. The proposed method has an ability to analyze both static and time-series data. When inferring a genetic network only from steady-state gene expression data, however, the proposed method is equivalent to GENIE3. Therefore, the proposed method can be seen as an extension of GENIE3. Through numerical experiments, we showed that the proposed method outperformed the existing inference methods on all of the 5 artificial genetic network inference problems.

MOTIVATION: Drug combination therapy for treatment of cancers and other multifactorial diseases has the potential of increasing the therapeutic effect, while reducing the likelihood of drug resistance. In order to reduce time and cost spent in comprehensive screens, methods are needed which can model additive effects of possible drug combinations. RESULTS: We here show that the transcriptional response to combinatorial drug treatment at promoters, as measured by single molecule CAGE technology, is accurately described by a linear combination of the responses of the individual drugs at a genome wide scale. We also find that the same linear relationship holds for transcription at enhancer elements. We conclude that the described approach is promising for eliciting the transcriptional response to multidrug treatment at promoters and enhancers in an unbiased genome wide way, which may minimize the need for exhaustive combinatorial screens. AVAILABILITY AND IMPLEMENTATION: The CAGE sequence data used in this study is available in the DDBJ Sequence Read Archive (http://trace.ddbj.nig.ac.jp/index_e.html), accession number DRP001113. CONTACT: [email protected] or [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

In order to analyze short time-series of expression levels of many genes, we can use some clustering technique to identify genes with similar expression patterns, group them together, and then infer a network consisting of the constructed clusters. But, it is inadequate to use the clustering technique before inferring a network. Instead, several researchers have proposed the methods that execute the gene clustering and the network inference simultaneously. As these methods use a linear model to describe networks, however, they are unsuitable to analyze time-series of gene expression data. In order to overcome the drawbacks of the existing approaches, this study proposes a new method that executes the gene clustering and the network inference simultaneously. As the proposed method uses a set of non-linear differential equations to describe networks, it has an ability to analyze time-series of gene expression data. As the clustering algorithm, we use the robust EM curve clustering algorithm proposed by Chamroukhi. Through numerical experiments, we finally show the effectiveness of the proposed method.

A number of genetic network inference methods have been proposed. These methods often infer many erroneous regulations. In order to decrease the number of erroneous regulations, this study uses a priori knowledge that biochemical networks exhibit hierarchical structures. This study detects the hierarchical structure in the target network using a hierarchical random graph model proposed by Clauset and colleagues. When the regulations inferred by the inference method are inconsistent with the detected hierarchical structure, we can conclude that they are unreasonable. However, it is not always easy to detect the hierarchical structure in the target network because of the regulations erroneously inferred by the inference method. In order to obtain a reasonable hierarchical structure, this study first infers a large number of genetic networks from the observed gene expression data by using a method that combines a genetic network inference method with a bootstrap method. We then extract a hierarchical structure from the inferred multiple genetic networks so that it is consistent with most of the networks. Through numerical experiments, we finally show that the use of the hierarchical structure in the network improves the reliabilities of regulations inferred by the genetic network inference method.

The immediate-early response mediates cell fate in response to a variety of extracellular stimuli and is dysregulated in many cancers. However, the specificity of the response across stimuli and cell types, and the roles of non-coding RNAs are not well understood. Using a large collection of densely-sampled time series expression data we have examined the induction of the immediate-early response in unparalleled detail, across cell types and stimuli. We exploit cap analysis of gene expression (CAGE) time series datasets to directly measure promoter activities over time. Using a novel analysis method for time series data we identify transcripts with expression patterns that closely resemble the dynamics of known immediate-early genes (IEGs) and this enables a comprehensive comparative study of these genes and their chromatin state. Surprisingly, these data suggest that the earliest transcriptional responses often involve promoters generating non-coding RNAs, many of which are produced in advance of canonical protein-coding IEGs. IEGs are known to be capable of induction without de novo protein synthesis. Consistent with this, we find that the response of both protein-coding and non-coding RNA IEGs can be explained by their transcriptionally poised, permissive chromatin state prior to stimulation. We also explore the function of non-coding RNAs in the attenuation of the immediate early response in a small RNA sequencing dataset matched to the CAGE data: We identify a novel set of microRNAs responsible for the attenuation of the IEG response in an estrogen receptor positive cancer cell line. Our computational statistical method is well suited to meta-analyses as there is no requirement for transcripts to pass thresholds for significant differential expression between time points, and it is agnostic to the number of time points per dataset.

Vohradský has proposed a neural network model to describe biochemical networks. Based on this model, several researchers have proposed genetic network inference methods. When trying to analyze large-scale genetic networks, however, these methods must solve high-dimensional function optimization problems. In order to resolve the high-dimensionality in the estimation of the parameters of the Vohradský's neural network model, this study proposes a new method. The proposed method estimates the parameters of the neural network model by solving two-dimensional function optimization problems. Although these two-dimensional problems are non-linear, their low-dimensionality would make the estimation of the model parameters easier. Finally, we confirm the effectiveness of the proposed method through numerical experiments.

For the inference of S-system models of genetic networks, this study proposes a new method, i.e., a two-phase estimation method. The two-phase estimation method is an extension of the decoupling approach proposed by Voit and Almeida. The decoupling approach defines the estimation of S-system parameters as a problem of solving sets of non-linear algebraic equations. Our method first transforms each set of non-linear algebraic equations, that is defined by the decoupling approach, into a set of linear ones. The transformation of the equations is easily accomplished by solving a linear programming problem. The proposed method then estimates S-system parameters by solving the transformed linear equations. As the proposed two-phase estimation method infers an S-system model only by solving linear programming problems and sets of linear algebraic equations, it always provides us with a unique solution. Moreover, its computational cost is very low. Finally, we confirm the effectiveness of the proposed method through numerical experiments.

MOTIVATION: Due to recent advances in high-throughput technologies, data on various types of genomic annotation have accumulated. These data will be crucially helpful for elucidating the combinatorial logic of transcription. Although several approaches have been proposed for inferring cooperativity among multiple factors, most approaches are haunted by the issues of normalization and threshold values. RESULTS: In this article, we propose a rank-based non-parametric statistical test for measuring the effects between two gene sets. This method is free from the issues of normalization and threshold value determination for gene expression values. Furthermore, we have proposed an efficient Markov chain Monte Carlo method for calculating an approximate significance value of synergy. We have applied this approach for detecting synergistic combinations of transcription factor binding motifs and histone modifications. AVAILABILITY: C implementation of the method is available from http://www.hgc.jp/~yshira/software/rankSynergy.zip. CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

An S-system model is considered as an ideal model for describing genetic networks. As one of effective techniques for inferring S-system models of genetic networks, the problem decomposition strategy has been proposed. This strategy defines the inference of a genetic network consisting of N genes as N subproblems, each of which is a 2(N+1)-dimensional function optimization problem. When we try to infer large-scale genetic networks consisting of many genes, however, it is not always easy for function optimization algorithms to solve 2(N + 1)-dimensional problems. In this study, we thus propose a new technique that transforms the 2(N + 1)-dimensional S-system parameter estimation problems into (N+2)-dimensional problems. The proposed technique reduces the search dimensions of the problems by solving linear programming problems. The transformed problems are then optimized using evolutionary algorithms. Finally, through numerical experiments on an artificial genetic network inference problem, we show that the proposed dimension reduction approach is more than 3 times faster than the problem decomposition approach.

MOTIVATION: Clustering and gene network inference often help to predict the biological functions of gene subsets. Recently, researchers have accumulated a large amount of time-course transcriptome data collected under different treatment conditions to understand the physiological states of cells in response to extracellular stimuli and to identify drug-responsive genes. Although a variety of statistical methods for clustering and inferring gene networks from expression profiles have been proposed, most of these are not tailored to simultaneously treat expression data collected under multiple stimulation conditions. RESULTS: We propose a new statistical method for analyzing temporal profiles under multiple experimental conditions. Our method simultaneously performs clustering of temporal expression profiles and inference of regulatory relationships among gene clusters. We applied this method to MCF7 human breast cancer cells treated with epidermal growth factor and heregulin which induce cellular proliferation and differentiation, respectively. The results showed that the method is useful for extracting biologically relevant information. AVAILABILITY: A MATLAB implementation of the method is available from http://csb.gsc.riken.jp/yshira/software/clusterNetwork.zip.

Recently, the inference of genetic networks was defined as a series of discrimination tasks. The inference method based on this problem definition infers genetic networks by obtaining predictors that can predict the signs of the differential coefficients of the gene expression levels. As these predictors are obtained by solving linear programming problems, the computational time of the method is very short. The method however has no explicit mechanism to utilize a priori knowledge about genetic networks. This study therefore extends the inference method based on the discrimination tasks to make it possible to utilize the a priori knowledge. In order to verify its effectiveness, we then apply the modified method to artificial genetic network inference problems.

MOTIVATION: Genetic network inference methods based on sets of differential equations generally require a great deal of time, as the equations must be solved many times. To reduce the computational cost, researchers have proposed other methods for inferring genetic networks by solving sets of differential equations only a few times, or even without solving them at all. When we try to obtain reasonable network models using these methods, however, we must estimate the time derivatives of the gene expression levels with great precision. In this study, we propose a new method to overcome the drawbacks of inference methods based on sets of differential equations. RESULTS: Our method infers genetic networks by obtaining classifiers capable of predicting the signs of the derivatives of the gene expression levels. For this purpose, we defined a genetic network inference problem as a series of discrimination tasks, then solved the defined series of discrimination tasks with a linear programming machine. Our experimental results demonstrated that the proposed method is capable of correctly inferring genetic networks, and doing so more than 500 times faster than the other inference methods based on sets of differential equations. Next, we applied our method to actual expression data of the bacterial SOS DNA repair system. And finally, we demonstrated that our approach relates to the inference method based on the S-system model. Though our method provides no estimation of the kinetic parameters, it should be useful for researchers interested only in the network structure of a target system. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: The inference of a genetic network is a problem in which mutual interactions among genes are deduced using time-series of gene expression patterns. While a number of models have been proposed to describe genetic regulatory networks, this study focuses on a set of differential equations since it has the ability to model dynamic behavior of gene expression. When we use a set of differential equations to describe genetic networks, the inference problem can be defined as a function approximation problem. On the basis of this problem definition, we propose in this study a new method to infer reduced NGnet models of genetic networks. RESULTS: Through numerical experiments on artificial genetic network inference problems, we demonstrated that our method has the ability to infer genetic networks correctly and it was faster than the other inference methods. We then applied the proposed method to actual expression data of the bacterial SOS DNA repair system, and succeeded in finding several reasonable regulations. When our method inferred the genetic network from the actual data, it required about 4.7 min on a single-CPU personal computer. CONCLUSION: The proposed method has an ability to obtain reasonable networks with a short computational time. As a high performance computer is not always available at every laboratory, the short computational time of our method is a preferable feature. There does not seem to be a perfect model for the inference of genetic networks yet. Therefore, in order to extract reliable information from the observed gene expression data, we should infer genetic networks using multiple inference methods based on different models. Our approach could be used as one of the promising inference methods.

The inference of genetic networks using a model based on a set of differential equations is generally time-consuming. In order to decrease its computational time, we have proposed the inference method using a normalized Gaussian network (NGnet) model. The inferred models however contain many false-positive regulations when we apply the NGnet approach to the genetic network inference problems. This paper proposes the reduced NGnet model and the gradual reduction strategy to overcome the drawbacks of the NGnet approach. Then, in order to verify their effectiveness, we apply the inference method using the proposed techniques to several artificial genetic network inference problems.

A model based on a set of differential equations can effectively capture various dynamics. This type of model is, therefore, ideal for describing genetic networks. The genetic network inference problem based on a set of differential equations is generally defined as a parameter estimation problem. On the basis of this problem definition, several computational methods have been proposed so far. On the other hand, the genetic network inference problem based on a set of differential equations can be also defined as a function approximation problem. For solving the defined function approximation problem, any type of function approximator is available. In this study, on the basis of the latter problem definition, we propose a new method for the inference of genetic networks using a normalized Gaussian network model. As the EM algorithm is available for the learning of the NGnet model, the computational time of the proposed method is much shorter than those of other inference methods. The effectiveness of the proposed inference method is verified through numerical experiments of several artificial genetic network inference problems.

We propose a new method for the inference of the genetic networks. The proposed method uses a neural network model to describe the genetic network. The inference of the neural network model of the genetic network is defined as the function optimization problem. As the function optimizer for this problem, a genetic local search is used. At this time, to enhance the probability of finding a reasonable solution, we introduce a priori knowledge about the genetic network into the objective function. In this paper, we also propose the method based on the sensitivity analysis to interpret the optimized neural network model. Through artificial genetic network inference problems, we verify the effectiveness of the proposed method.

MOTIVATION: To resolve the high-dimensionality of the genetic network inference problem in the S-system model, a problem decomposition strategy has been proposed. While this strategy certainly shows promise, it cannot provide a model readily applicable to the computational simulation of the genetic network when the given time-series data contain measurement noise. This is a significant limitation of the problem decomposition, given that our analysis and understanding of the genetic network depend on the computational simulation. RESULTS: We propose a new method for inferring S-system models of large-scale genetic networks. The proposed method is based on the problem decomposition strategy and a cooperative coevolutionary algorithm. As the subproblems divided by the problem decomposition strategy are solved simultaneously using the cooperative coevolutionary algorithm, the proposed method can be used to infer any S-system model ready for computational simulation. To verify the effectiveness of the proposed method, we apply it to two artificial genetic network inference problems. Finally, the proposed method is used to analyze the actual DNA microarray data.

UNLABELLED: OBIYagns (yet another gene network simulator) is a biochemical system simulator that comprises a multiple-user Web-based graphical interface, an ordinary differential equation solver and a parameter estimators distributed over an open bioinformatics grid (OBIGrid). This grid-based biochemical simulation system can achieve high performance and provide a secure simulation environment for estimating kinetic parameters in an acceptable time period. OBIYagns can be applied to larger system biology-oriented simulation projects. AVAILABILITY: OBIYagns example models, methods and user guide are available at https://access.obigrid.org/yagns/ SUPPLEMENTARY INFORMATION: Please refer to Bioinformatics online.

In this paper, we propose a new method for the inference of S-system models of large-scale genetic networks. This method employs a technique to decompose the genetic network inference problem into several subproblems, and then applies a genetic local search to each of the subproblems. A local search method utilizing the feature of the S-system model is used as one of the search operators in this genetic local search. Finally, the effectiveness of the proposed method is verified through a genetic network inference problem.