A Bayesian approach to joint modeling of protein-DNA binding, gene expression and sequence data

A systematic review of state-of-the-art strategies for machine learning-based protein function prediction

New drug discovery is inseparable from the discovery of drug targets, and the vast majority of the known targets are proteins. At the same time, proteins are essential structural and functional elements of living cells necessary for the maintenance of all forms of life. Therefore, protein functions have become the focus of many pharmacological and biological studies. Traditional experimental techniques are no longer adequate for rapidly growing annotation of protein sequences, and approaches to protein function prediction using computational methods have emerged and flourished. A significant trend has been to use machine learning to achieve this goal. In this review, approaches to protein function prediction based on the sequence, structure, protein-protein interaction (PPI) networks, and fusion of multi-information sources are discussed. The current status of research on protein function prediction using machine learning is considered, and existing challenges and prominent breakthroughs are discussed to provide ideas and methods for future studies.

Semi-Supervised Learning Using Hierarchical Mixture Models: Gene Essentiality Case Study

Integrating gene-level data is useful for predicting the role of genes in biological processes. This problem has typically focused on supervised classification, which requires large training sets of positive and negative examples. However, training data sets that are too small for supervised approaches can still provide valuable information. We describe a hierarchical mixture model that uses limited positively labeled gene training data for semi-supervised learning. We focus on the problem of predicting essential genes, where a gene is required for the survival of an organism under particular conditions. We applied cross-validation and found that the inclusion of positively labeled samples in a semi-supervised learning framework with the hierarchical mixture model improves the detection of essential genes compared to unsupervised, supervised, and other semi-supervised approaches. There was also improved prediction performance when genes are incorrectly assumed to be non-essential. Our comparisons indicate that the incorporation of even small amounts of existing knowledge improves the accuracy of prediction and decreases variability in predictions. Although we focused on gene essentiality, the hierarchical mixture model and semi-supervised framework is standard for problems focused on prediction of genes or other features, with multiple data types characterizing the feature, and a small set of positive labels.

Bayesian state-space modeling in gene expression data analysis: An application with biomarker prediction

Background and ObjectiveBayesian State Space models are recent advancement in stochastic modeling which capture the randomness of a hidden background process by scrutinizing the prior knowledge and likelihood of observed data. This article elucidate the scope of Bayesian state space modeling on predicting the future expression values of a longitudinal micro array data. MethodsThe study conveniently makes use of longitudinally collected clinical trial data (GSE30531) from NCBI Gene Expression Omnibus (GEO) data repository. Multiple testing methodology using t-test is used for selecting differentially expressed genes between groups for fitting the model. The parameter values of the predictive model and future expression levels are estimated by drawing samples from the posterior joint distribution using a stochastic Markov Chain Monte Carlo (MCMC) algorithm which relies on Gibbs Sampling. The study also made an attempt to get estimates and its 95% Credible Interval through assumptions of different covariance structures like Variance Components, First order Auto Regressive and Unstructured variance-covariance structure to showcase the flexibility of the algorithm. Results72 Distinct genes with significantly different expression levels where selected for model fitting. Parameter estimates showed almost similar trends under different covariance structure assumption. Cross tabulation of gene frequencies having minimum credible interval under each covariance structure and study group showed a significant P value of 0.02. ConclusionsPresent study reveals that Bayesian state space models can be effectively used to explain and predict a complex data like gene expression data.

A graphical model method for integrating multiple sources of genome-scale data

Making effective use of multiple data sources is a major challenge in modern bioinformatics. Genome-wide data such as measures of transcription factor binding, gene expression, and sequence conservation, which are used to identify binding regions and genes that are important to major biological processes such as development and disease, can be difficult to use together due to the different biological meanings and statistical distributions of the heterogeneous data types, but each can provide valuable information for understanding the processes under study. Here we present methods for integrating multiple data sources to gain a more complete picture of gene regulation and expression. Our goal is to identify genes and cis-regulatory regions which play specific biological roles. We describe a graphical mixture model approach for data integration, examine the effect of using different model topologies, and discuss methods for evaluating the effectiveness of the models. Model fitting is computationally efficient and produces results which have clear biological and statistical interpretations. The Hedgehog and Dorsal signaling pathways in Drosophila, which are critical in embryonic development, are used as examples.

Bayesian hidden Markov models to identify RNA-protein interaction sites in PAR-CLIP

The photoactivatable ribonucleoside enhanced cross-linking immunoprecipitation (PAR-CLIP) has been increasingly used for the global mapping of RNA-protein interaction sites. There are two key features of the PAR-CLIP experiments: The sequence read tags are likely to form an enriched peak around each RNA-protein interaction site; and the cross-linking procedure is likely to introduce a specific mutation in each sequence read tag at the interaction site. Several ad hoc methods have been developed to identify the RNA-protein interaction sites using either sequence read counts or mutation counts alone; however, rigorous statistical methods for analyzing PAR-CLIP are still lacking. In this article, we propose an integrative model to establish a joint distribution of observed read and mutation counts. To pinpoint the interaction sites at single base-pair resolution, we developed a novel modeling approach that adopts non-homogeneous hidden Markov models to incorporate the nucleotide sequence at each genomic location. Both simulation studies and data application showed that our method outperforms the ad hoc methods, and provides reliable inferences for the RNA-protein binding sites from PAR-CLIP data.

Detection of candidate tumor driver genes using a fully integrated Bayesian approach

DNA copy number alterations (CNAs), including amplifications and deletions, can result in significant changes in gene expression and are closely related to the development and progression of many diseases, especially cancer. For example, CNA-associated expression changes in certain genes (called candidate tumor driver genes) can alter the expression levels of many downstream genes through transcription regulation and cause cancer. Identification of such candidate tumor driver genes leads to discovery of novel therapeutic targets for personalized treatment of cancers. Several approaches have been developed for this purpose by using both copy number and gene expression data. In this study, we propose a Bayesian approach to identify candidate tumor driver genes, in which the copy number and gene expression data are modeled together, and the dependency between the two data types is modeled through conditional probabilities. The proposed joint modeling approach can identify CNA and differentially expressed genes simultaneously, leading to improved detection of candidate tumor driver genes and comprehensive understanding of underlying biological processes. We evaluated the proposed method in simulation studies, and then applied to a head and neck squamous cell carcinoma data set. Both simulation studies and data application show that the joint modeling approach can significantly improve the performance in identifying candidate tumor driver genes, when compared with other existing approaches.

Inferring functional transcription factor-gene binding pairs by integrating transcription factor binding data with transcription factor knockout data

Background

Chromatin immunoprecipitation (ChIP) experiments are now the most comprehensive experimental approaches for mapping the binding of transcription factors (TFs) to their target genes. However, ChIP data alone is insufficient for identifying functional binding target genes of TFs for two reasons. First, there is an inherent high false positive/negative rate in ChIP-chip or ChIP-seq experiments. Second, binding signals in the ChIP data do not necessarily imply functionality.

Methods

It is known that ChIP-chip data and TF knockout (TFKO) data reveal complementary information on gene regulation. While ChIP-chip data can provide TF-gene binding pairs, TFKO data can provide TF-gene regulation pairs. Therefore, we propose a novel network approach for identifying functional TF-gene binding pairs by integrating the ChIP-chip data with the TFKO data. In our method, a TF-gene binding pair from the ChIP-chip data is regarded to be functional if it also has high confident curated TFKO TF-gene regulatory relation or deduced hypostatic TF-gene regulatory relation.

Results and conclusions

We first validated our method on a gathered ground truth set. Then we applied our method to the ChIP-chip data to identify functional TF-gene binding pairs. The biological significance of our identified functional TF-gene binding pairs was shown by assessing their functional enrichment, the prevalence of protein-protein interaction, and expression coherence. Our results outperformed the results of three existing methods across all measures. And our identified functional targets of TFs also showed statistical significance over the randomly assigned TF-gene pairs. We also showed that our method is dataset independent and can apply to ChIP-seq data and the E. coli genome. Finally, we provided an example showing the biological applicability of our notion.

Bayesian Joint Modeling of Multiple Gene Networks and Diverse Genomic Data to Identify Target Genes of a Transcription Factor

We consider integrative modeling of multiple gene networks and diverse genomic data, including protein-DNA binding, gene expression and DNA sequence data, to accurately identify the regulatory target genes of a transcription factor (TF). Rather than treating all the genes equally and independently a priori in existing joint modeling approaches, we incorporate the biological prior knowledge that neighboring genes on a gene network tend to be (or not to be) regulated together by a TF. A key contribution of our work is that, to maximize the use of all existing biological knowledge, we allow incorporation of multiple gene networks into joint modeling of genomic data by introducing a mixture model based on the use of multiple Markov random fields (MRFs). Another important contribution of our work is to allow different genomic data to be correlated and to examine the validity and effect of the independence assumption as adopted in existing methods. Due to a fully Bayesian approach, inference about model parameters can be carried out based on MCMC samples. Application to an E. coli data set, together with simulation studies, demonstrates the utility and statistical efficiency gains with the proposed joint model.

Statistical methods for integrating multiple types of high-throughput data

Large-scale sequencing, copy number, mRNA, and protein data have given great promise to the biomedical research, while posing great challenges to data management and data analysis. Integrating different types of high-throughput data from diverse sources can increase the statistical power of data analysis and provide deeper biological understanding. This chapter uses two biomedical research examples to illustrate why there is an urgent need to develop reliable and robust methods for integrating the heterogeneous data. We then introduce and review some recently developed statistical methods for integrative analysis for both statistical inference and classification purposes. Finally, we present some useful public access databases and program code to facilitate the integrative analysis in practice.