An integrated probabilistic model for functional prediction of proteins

Are trait-associated genes clustered together in a gene network?

Genome-wide association studies (GWAS) have provided an abundance of information about the genetic variants and their loci that are associated to complex traits and diseases. However, due to linkage disequilibrium (LD) and noncoding regions of loci, it remains a challenge to pinpoint the causal genes. Gene network-based approaches, paired with network diffusion methods, have been proposed to prioritize causal genes and to boost statistical power in GWAS based on the assumption that trait-associated genes are clustered in a gene network. Due to the difficulty in mapping trait-associated variants to genes in GWAS, this assumption has never been directly or rigorously tested empirically. On the other hand, whole exome sequencing (WES) data focuses on the protein-coding regions, directly identifying trait-associated genes. In this study, we tested the assumption by leveraging the recently available exome-based association statistics from the UK Biobank WES data along with two types of networks. We found that almost all trait-associated genes were significantly more proximal to each other than randomly selected genes within both networks. These results support the assumption that trait-associated genes are clustered in gene networks, which can be further leveraged to boost the power of GWAS such as by introducing less stringent p value thresholds.

A Gated Recurrent Unit based architecture for recognizing ontology concepts from biological literature

Background

Annotating scientific literature with ontology concepts is a critical task in biology and several other domains for knowledge discovery. Ontology based annotations can power large-scale comparative analyses in a wide range of applications ranging from evolutionary phenotypes to rare human diseases to the study of protein functions. Computational methods that can tag scientific text with ontology terms have included lexical/syntactic methods, traditional machine learning, and most recently, deep learning.

Results

Here, we present state of the art deep learning architectures based on Gated Recurrent Units for annotating text with ontology concepts. We use the Colorado Richly Annotated Full Text Corpus (CRAFT) as a gold standard for training and testing. We explore a number of additional information sources including NCBI’s BioThesauraus and Unified Medical Language System (UMLS) to augment information from CRAFT for increasing prediction accuracy. Our best model results in a 0.84 F1 and semantic similarity.

Conclusion

The results shown here underscore the impact for using deep learning architectures for automatically recognizing ontology concepts from literature. The augmentation of the models with biological information beyond that present in the gold standard corpus shows a distinct improvement in prediction accuracy.

Integration of probabilistic functional networks without an external Gold Standard

BACKGROUND

Probabilistic functional integrated networks (PFINs) are designed to aid our understanding of cellular biology and can be used to generate testable hypotheses about protein function. PFINs are generally created by scoring the quality of interaction datasets against a Gold Standard dataset, usually chosen from a separate high-quality data source, prior to their integration. Use of an external Gold Standard has several drawbacks, including data redundancy, data loss and the need for identifier mapping, which can complicate the network build and impact on PFIN performance. Additionally, there typically are no Gold Standard data for non-model organisms.

RESULTS

We describe the development of an integration technique, ssNet, that scores and integrates both high-throughput and low-throughout data from a single source database in a consistent manner without the need for an external Gold Standard dataset. Using data from Saccharomyces cerevisiae we show that ssNet is easier and faster, overcoming the challenges of data redundancy, Gold Standard bias and ID mapping. In addition ssNet results in less loss of data and produces a more complete network.

CONCLUSIONS

The ssNet method allows PFINs to be built successfully from a single database, while producing comparable network performance to networks scored using an external Gold Standard source and with reduced data loss.

GenePlexus: a web-server for gene discovery using network-based machine learning

Biomedical researchers take advantage of high-throughput, high-coverage technologies to routinely generate sets of genes of interest across a wide range of biological conditions. Although these technologies have directly shed light on the molecular underpinnings of various biological processes and diseases, the list of genes from any individual experiment is often noisy and incomplete. Additionally, interpreting these lists of genes can be challenging in terms of how they are related to each other and to other genes in the genome. In this work, we present GenePlexus (https://www.geneplexus.net/), a web-server that allows a researcher to utilize a powerful, network-based machine learning method to gain insights into their gene set of interest and additional functionally similar genes. Once a user uploads their own set of human genes and chooses between a number of different human network representations, GenePlexus provides predictions of how associated every gene in the network is to the input set. The web-server also provides interpretability through network visualization and comparison to other machine learning models trained on thousands of known process/pathway and disease gene sets. GenePlexus is free and open to all users without the need for registration.

Interpretable network propagation with application to expanding the repertoire of human proteins that interact with SARS-CoV-2

BACKGROUND

Network propagation has been widely used for nearly 20 years to predict gene functions and phenotypes. Despite the popularity of this approach, little attention has been paid to the question of provenance tracing in this context, e.g., determining how much any experimental observation in the input contributes to the score of every prediction.

RESULTS

We design a network propagation framework with 2 novel components and apply it to predict human proteins that directly or indirectly interact with SARS-CoV-2 proteins. First, we trace the provenance of each prediction to its experimentally validated sources, which in our case are human proteins experimentally determined to interact with viral proteins. Second, we design a technique that helps to reduce the manual adjustment of parameters by users. We find that for every top-ranking prediction, the highest contribution to its score arises from a direct neighbor in a human protein-protein interaction network. We further analyze these results to develop functional insights on SARS-CoV-2 that expand on known biology such as the connection between endoplasmic reticulum stress, HSPA5, and anti-clotting agents.

CONCLUSIONS

We examine how our provenance-tracing method can be generalized to a broad class of network-based algorithms. We provide a useful resource for the SARS-CoV-2 community that implicates many previously undocumented proteins with putative functional relationships to viral infection. This resource includes potential drugs that can be opportunistically repositioned to target these proteins. We also discuss how our overall framework can be extended to other, newly emerging viruses.

Protein function prediction with gene ontology: from traditional to deep learning models

Protein function prediction is a crucial part of genome annotation. Prediction methods have recently witnessed rapid development, owing to the emergence of high-throughput sequencing technologies. Among the available databases for identifying protein function terms, Gene Ontology (GO) is an important resource that describes the functional properties of proteins. Researchers are employing various approaches to efficiently predict the GO terms. Meanwhile, deep learning, a fast-evolving discipline in data-driven approach, exhibits impressive potential with respect to assigning GO terms to amino acid sequences. Herein, we reviewed the currently available computational GO annotation methods for proteins, ranging from conventional to deep learning approach. Further, we selected some suitable predictors from among the reviewed tools and conducted a mini comparison of their performance using a worldwide challenge dataset. Finally, we discussed the remaining major challenges in the field, and emphasized the future directions for protein function prediction with GO.

Supervised learning is an accurate method for network-based gene classification

Background

Assigning every human gene to specific functions, diseases and traits is a grand challenge in modern genetics. Key to addressing this challenge are computational methods, such as supervised learning and label propagation, that can leverage molecular interaction networks to predict gene attributes. In spite of being a popular machine-learning technique across fields, supervised learning has been applied only in a few network-based studies for predicting pathway-, phenotype- or disease-associated genes. It is unknown how supervised learning broadly performs across different networks and diverse gene classification tasks, and how it compares to label propagation, the widely benchmarked canonical approach for this problem.

Results

In this study, we present a comprehensive benchmarking of supervised learning for network-based gene classification, evaluating this approach and a classic label propagation technique on hundreds of diverse prediction tasks and multiple networks using stringent evaluation schemes. We demonstrate that supervised learning on a gene’s full network connectivity outperforms label propagaton and achieves high prediction accuracy by efficiently capturing local network properties, rivaling label propagation’s appeal for naturally using network topology. We further show that supervised learning on the full network is also superior to learning on node embeddings (derived using node2vec), an increasingly popular approach for concisely representing network connectivity. These results show that supervised learning is an accurate approach for prioritizing genes associated with diverse functions, diseases and traits and should be considered a staple of network-based gene classification workflows.

Availability and implementation

The datasets and the code used to reproduce the results and add new gene classification methods have been made freely available.

Contact

arjun@msu.edu

Supplementary information

Supplementary data are available at Bioinformatics online.

Supervised learning is an accurate method for network-based gene classification

BACKGROUND

Assigning every human gene to specific functions, diseases and traits is a grand challenge in modern genetics. Key to addressing this challenge are computational methods, such as supervised learning and label propagation, that can leverage molecular interaction networks to predict gene attributes. In spite of being a popular machine-learning technique across fields, supervised learning has been applied only in a few network-based studies for predicting pathway-, phenotype- or disease-associated genes. It is unknown how supervised learning broadly performs across different networks and diverse gene classification tasks, and how it compares to label propagation, the widely benchmarked canonical approach for this problem.

RESULTS

In this study, we present a comprehensive benchmarking of supervised learning for network-based gene classification, evaluating this approach and a classic label propagation technique on hundreds of diverse prediction tasks and multiple networks using stringent evaluation schemes. We demonstrate that supervised learning on a gene's full network connectivity outperforms label propagaton and achieves high prediction accuracy by efficiently capturing local network properties, rivaling label propagation's appeal for naturally using network topology. We further show that supervised learning on the full network is also superior to learning on node embeddings (derived using node2vec), an increasingly popular approach for concisely representing network connectivity. These results show that supervised learning is an accurate approach for prioritizing genes associated with diverse functions, diseases and traits and should be considered a staple of network-based gene classification workflows.

AVAILABILITY AND IMPLEMENTATION

The datasets and the code used to reproduce the results and add new gene classification methods have been made freely available.

CONTACT

arjun@msu.edu.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Multitask Protein Function Prediction through Task Dissimilarity

Automated protein function prediction is a challenging problem with distinctive features, such as the hierarchical organization of protein functions and the scarcity of annotated proteins for most biological functions. We propose a multitask learning algorithm addressing both issues. Unlike standard multitask algorithms, which use task (protein functions) similarity information as a bias to speed up learning, we show that dissimilarity information enforces separation of rare class labels from frequent class labels, and for this reason is better suited for solving unbalanced protein function prediction problems. We support our claim by showing that a multitask extension of the label propagation algorithm empirically works best when the task relatedness information is represented using a dissimilarity matrix as opposed to a similarity matrix. Moreover, the experimental comparison carried out on three model organism shows that our method has a more stable performance in both "protein-centric" and "function-centric" evaluation settings.

Effusion: prediction of protein function from sequence similarity networks

Motivation

Critical evaluation of methods for protein function prediction shows that data integration improves the performance of methods that predict protein function, but a basic BLAST-based method is still a top contender. We sought to engineer a method that modernizes the classical approach while avoiding pitfalls common to state-of-the-art methods.

Results

We present a method for predicting protein function, Effusion, which uses a sequence similarity network to add context for homology transfer, a probabilistic model to account for the uncertainty in labels and function propagation, and the structure of the Gene Ontology (GO) to best utilize sparse input labels and make consistent output predictions. Effusion's model makes it practical to integrate rare experimental data and abundant primary sequence and sequence similarity. We demonstrate Effusion's performance using a critical evaluation method and provide an in-depth analysis. We also dissect the design decisions we used to address challenges for predicting protein function. Finally, we propose directions in which the framework of the method can be modified for additional predictive power.

Availability and implementation

The source code for an implementation of Effusion is freely available at https://github.com/babbittlab/effusion.

Supplementary information

Supplementary data are available at Bioinformatics online.

Predicting Protein Functions by Using Unbalanced Random Walk Algorithm on Three Biological Networks

With the gap between the sequence data and their functional annotations becomes increasing wider, many computational methods have been proposed to annotate functions for unknown proteins. However, designing effective methods to make good use of various biological resources is still a big challenge for researchers due to function diversity of proteins. In this work, we propose a new method named ThrRW, which takes several steps of random walking on three different biological networks: protein interaction network (PIN), domain co-occurrence network (DCN), and functional interrelationship network (FIN), respectively, so as to infer functional information from neighbors in the corresponding networks. With respect to the topological and structural differences of the three networks, the number of walking steps in the three networks will be different. In the course of working, the functional information will be transferred from one network to another according to the associations between the nodes in different networks. The results of experiment on S. cerevisiae data show that our method achieves better prediction performance not only than the methods that consider both PIN data and GO term similarities, but also than the methods using both PIN data and protein domain information, which verifies the effectiveness of our method on integrating multiple biological data sources.

An unsupervised learning approach to find ovarian cancer genes through integration of biological data

Cancer is a disease characterized largely by the accumulation of out-of-control somatic mutations during the lifetime of a patient. Distinguishing driver mutations from passenger mutations has posed a challenge in modern cancer research. With the advanced development of microarray experiments and clinical studies, a large numbers of candidate cancer genes have been extracted and distinguishing informative genes out of them is essential. As a matter of fact, we proposed to find the informative genes for cancer by using mutation data from ovarian cancers in our framework. In our model we utilized the patient gene mutation profile, gene expression data and gene gene interactions network to construct a graphical representation of genes and patients. Markov processes for mutation and patients are triggered separately. After this process, cancer genes are prioritized automatically by examining their scores at their stationary distributions in the eigenvector. Extensive experiments demonstrate that the integration of heterogeneous sources of information is essential in finding important cancer genes.

The clustering of functionally related genes contributes to CNV-mediated disease

Clusters of functionally related genes can be disrupted by a single copy number variant (CNV). We demonstrate that the simultaneous disruption of multiple functionally related genes is a frequent and significant characteristic of de novo CNVs in patients with developmental disorders (P = 1 × 10⁻³). Using three different functional networks, we identified unexpectedly large numbers of functionally related genes within de novo CNVs from two large independent cohorts of individuals with developmental disorders. The presence of multiple functionally related genes was a significant predictor of a CNV's pathogenicity when compared to CNVs from apparently healthy individuals and a better predictor than the presence of known disease or haploinsufficient genes for larger CNVs. The functionally related genes found in the de novo CNVs belonged to 70% of all clusters of functionally related genes found across the genome. De novo CNVs were more likely to affect functional clusters and affect them to a greater extent than benign CNVs (P = 6 × 10⁻⁴). Furthermore, such clusters of functionally related genes are phenotypically informative: Different patients possessing CNVs that affect the same cluster of functionally related genes exhibit more similar phenotypes than expected (P < 0.05). The spanning of multiple functionally similar genes by single CNVs contributes substantially to how these variants exert their pathogenic effects.

A survey of computational intelligence techniques in protein function prediction

During the past, there was a massive growth of knowledge of unknown proteins with the advancement of high throughput microarray technologies. Protein function prediction is the most challenging problem in bioinformatics. In the past, the homology based approaches were used to predict the protein function, but they failed when a new protein was different from the previous one. Therefore, to alleviate the problems associated with homology based traditional approaches, numerous computational intelligence techniques have been proposed in the recent past. This paper presents a state-of-the-art comprehensive review of various computational intelligence techniques for protein function predictions using sequence, structure, protein-protein interaction network, and gene expression data used in wide areas of applications such as prediction of DNA and RNA binding sites, subcellular localization, enzyme functions, signal peptides, catalytic residues, nuclear/G-protein coupled receptors, membrane proteins, and pathway analysis from gene expression datasets. This paper also summarizes the result obtained by many researchers to solve these problems by using computational intelligence techniques with appropriate datasets to improve the prediction performance. The summary shows that ensemble classifiers and integration of multiple heterogeneous data are useful for protein function prediction.

Disease gene identification by using graph kernels and Markov random fields

Genes associated with similar diseases are often functionally related. This principle is largely supported by many biological data sources, such as disease phenotype similarities, protein complexes, protein-protein interactions, pathways and gene expression profiles. Integrating multiple types of biological data is an effective method to identify disease genes for many genetic diseases. To capture the gene-disease associations based on biological networks, a kernel-based MRF method is proposed by combining graph kernels and the Markov random field (MRF) method. In the proposed method, three kinds of kernels are employed to describe the overall relationships of vertices in five biological networks, respectively, and a novel weighted MRF method is developed to integrate those data. In addition, an improved Gibbs sampling procedure and a novel parameter estimation method are proposed to generate predictions from the kernel-based MRF method. Numerical experiments are carried out by integrating known gene-disease associations, protein complexes, protein-protein interactions, pathways and gene expression profiles. The proposed kernel-based MRF method is evaluated by the leave-one-out cross validation paradigm, achieving an AUC score of 0.771 when integrating all those biological data in our experiments, which indicates that our proposed method is very promising compared with many existing methods.

Identifying disease genes by integrating multiple data sources

BACKGROUND

Now multiple types of data are available for identifying disease genes. Those data include gene-disease associations, disease phenotype similarities, protein-protein interactions, pathways, gene expression profiles, etc.. It is believed that integrating different kinds of biological data is an effective method to identify disease genes.

RESULTS

In this paper, we propose a multiple data integration method based on the theory of Markov random field (MRF) and the method of Bayesian analysis for identifying human disease genes. The proposed method is not only flexible in easily incorporating different kinds of data, but also reliable in predicting candidate disease genes.

CONCLUSIONS

Numerical experiments are carried out by integrating known gene-disease associations, protein complexes, protein-protein interactions, pathways and gene expression profiles. Predictions are evaluated by the leave-one-out method. The proposed method achieves an AUC score of 0.743 when integrating all those biological data in our experiments.

Using biological networks to improve our understanding of infectious diseases

Infectious diseases are the leading cause of death, particularly in developing countries. Although many drugs are available for treating the most common infectious diseases, in many cases the mechanism of action of these drugs or even their targets in the pathogen remain unknown. In addition, the key factors or processes in pathogens that facilitate infection and disease progression are often not well understood. Since proteins do not work in isolation, understanding biological systems requires a better understanding of the interconnectivity between proteins in different pathways and processes, which includes both physical and other functional interactions. Such biological networks can be generated within organisms or between organisms sharing a common environment using experimental data and computational predictions. Though different data sources provide different levels of accuracy, confidence in interactions can be measured using interaction scores. Connections between interacting proteins in biological networks can be represented as graphs and edges, and thus studied using existing algorithms and tools from graph theory. There are many different applications of biological networks, and here we discuss three such applications, specifically applied to the infectious disease tuberculosis, with its causative agent Mycobacterium tuberculosis and host, Homo sapiens. The applications include the use of the networks for function prediction, comparison of networks for evolutionary studies, and the generation and use of host–pathogen interaction networks.

Integrative Pathway Analysis Using Graph-Based Learning with Applications to TCGA Colon and Ovarian Data

Recent method development has included multi-dimensional genomic data algorithms because such methods have more accurately predicted clinical phenotypes related to disease. This study is the first to conduct an integrative genomic pathway-based analysis with a graph-based learning algorithm. The methodology of this analysis, graph-based semi-supervised learning, detects pathways that improve prediction of a dichotomous variable, which in this study is cancer stage. This analysis integrates genome-level gene expression, methylation, and single nucleotide polymorphism (SNP) data in serous cystadenocarcinoma (OV) and colon adenocarcinoma (COAD). The top 10 ranked predictive pathways in COAD and OV were biologically relevant to their respective cancer stages and significantly enhanced prediction accuracy and area under the ROC curve (AUC) when compared to single data-type analyses. This method is an effective way to simultaneously predict binary clinical phenotypes and discover their biological mechanisms.

An extensive analysis of disease-gene associations using network integration and fast kernel-based gene prioritization methods

OBJECTIVE

In the context of "network medicine", gene prioritization methods represent one of the main tools to discover candidate disease genes by exploiting the large amount of data covering different types of functional relationships between genes. Several works proposed to integrate multiple sources of data to improve disease gene prioritization, but to our knowledge no systematic studies focused on the quantitative evaluation of the impact of network integration on gene prioritization. In this paper, we aim at providing an extensive analysis of gene-disease associations not limited to genetic disorders, and a systematic comparison of different network integration methods for gene prioritization.

MATERIALS AND METHODS

We collected nine different functional networks representing different functional relationships between genes, and we combined them through both unweighted and weighted network integration methods. We then prioritized genes with respect to each of the considered 708 medical subject headings (MeSH) diseases by applying classical guilt-by-association, random walk and random walk with restart algorithms, and the recently proposed kernelized score functions.

RESULTS

The results obtained with classical random walk algorithms and the best single network achieved an average area under the curve (AUC) across the 708 MeSH diseases of about 0.82, while kernelized score functions and network integration boosted the average AUC to about 0.89. Weighted integration, by exploiting the different "informativeness" embedded in different functional networks, outperforms unweighted integration at 0.01 significance level, according to the Wilcoxon signed rank sum test. For each MeSH disease we provide the top-ranked unannotated candidate genes, available for further bio-medical investigation.

CONCLUSIONS

Network integration is necessary to boost the performances of gene prioritization methods. Moreover the methods based on kernelized score functions can further enhance disease gene ranking results, by adopting both local and global learning strategies, able to exploit the overall topology of the network.

Hierarchical ensemble methods for protein function prediction

Protein function prediction is a complex multiclass multilabel classification problem, characterized by multiple issues such as the incompleteness of the available annotations, the integration of multiple sources of high dimensional biomolecular data, the unbalance of several functional classes, and the difficulty of univocally determining negative examples. Moreover, the hierarchical relationships between functional classes that characterize both the Gene Ontology and FunCat taxonomies motivate the development of hierarchy-aware prediction methods that showed significantly better performances than hierarchical-unaware "flat" prediction methods. In this paper, we provide a comprehensive review of hierarchical methods for protein function prediction based on ensembles of learning machines. According to this general approach, a separate learning machine is trained to learn a specific functional term and then the resulting predictions are assembled in a "consensus" ensemble decision, taking into account the hierarchical relationships between classes. The main hierarchical ensemble methods proposed in the literature are discussed in the context of existing computational methods for protein function prediction, highlighting their characteristics, advantages, and limitations. Open problems of this exciting research area of computational biology are finally considered, outlining novel perspectives for future research.

Predicting gene ontology annotations of orphan GWAS genes using protein-protein interactions

Background

The number of genome-wide association studies (GWAS) has increased rapidly in the past couple of years, resulting in the identification of genes associated with different diseases. The next step in translating these findings into biomedically useful information is to find out the mechanism of the action of these genes. However, GWAS studies often implicate genes whose functions are currently unknown; for example, MYEOV, ANKLE1, TMEM45B and ORAOV1 are found to be associated with breast cancer, but their molecular function is unknown.

Results

We carried out Bayesian inference of Gene Ontology (GO) term annotations of genes by employing the directed acyclic graph structure of GO and the network of protein-protein interactions (PPIs). The approach is designed based on the fact that two proteins that interact biophysically would be in physical proximity of each other, would possess complementary molecular function, and play role in related biological processes. Predicted GO terms were ranked according to their relative association scores and the approach was evaluated quantitatively by plotting the precision versus recall values and F-scores (the harmonic mean of precision and recall) versus varying thresholds. Precisions of ~58% and ~ 40% for localization and functions respectively of proteins were determined at a threshold of ~30 (top 30 GO terms in the ranked list). Comparison with function prediction based on semantic similarity among nodes in an ontology and incorporation of those similarities in a k-nearest neighbor classifier confirmed that our results compared favorably.

Conclusions

This approach was applied to predict the cellular component and molecular function GO terms of all human proteins that have interacting partners possessing at least one known GO annotation. The list of predictions is available at http://severus.dbmi.pitt.edu/engo/GOPRED.html. We present the algorithm, evaluations and the results of the computational predictions, especially for genes identified in GWAS studies to be associated with diseases, which are of translational interest.

Improving protein function prediction using domain and protein complexes in PPI networks

Background

Characterization of unknown proteins through computational approaches is one of the most challenging problems in silico biology, which has attracted world-wide interests and great efforts. There have been some computational methods proposed to address this problem, which are either based on homology mapping or in the context of protein interaction networks.

Results

In this paper, two algorithms are proposed by integrating the protein-protein interaction (PPI) network, proteins’ domain information and protein complexes. The one is domain combination similarity (DCS), which combines the domain compositions of both proteins and their neighbors. The other is domain combination similarity in context of protein complexes (DSCP), which extends the protein functional similarity definition of DCS by combining the domain compositions of both proteins and the complexes including them. The new algorithms are tested on networks of the model species of Saccharomyces cerevisiae to predict functions of unknown proteins using cross validations. Comparing with other several existing algorithms, the results have demonstrated the effectiveness of our proposed methods in protein function prediction. Furthermore, the algorithm DSCP using experimental determined complex data is robust when a large percentage of the proteins in the network is unknown, and it outperforms DCS and other several existing algorithms.

Conclusions

The accuracy of predicting protein function can be improved by integrating the protein-protein interaction (PPI) network, proteins’ domain information and protein complexes.

Novel cardiovascular gene functions revealed via systematic phenotype prediction in zebrafish

Comprehensive functional annotation of vertebrate genomes is fundamental to biological discovery. Reverse genetic screening has been highly useful for determination of gene function, but is untenable as a systematic approach in vertebrate model organisms given the number of surveyable genes and observable phenotypes. Unbiased prediction of gene-phenotype relationships offers a strategy to direct finite experimental resources towards likely phenotypes, thus maximizing de novo discovery of gene functions. Here we prioritized genes for phenotypic assay in zebrafish through machine learning, predicting the effect of loss of function of each of 15,106 zebrafish genes on 338 distinct embryonic anatomical processes. Focusing on cardiovascular phenotypes, the learning procedure predicted known knockdown and mutant phenotypes with high precision. In proof-of-concept studies we validated 16 high-confidence cardiac predictions using targeted morpholino knockdown and initial blinded phenotyping in embryonic zebrafish, confirming a significant enrichment for cardiac phenotypes as compared with morpholino controls. Subsequent detailed analyses of cardiac function confirmed these results, identifying novel physiological defects for 11 tested genes. Among these we identified tmem88a, a recently described attenuator of Wnt signaling, as a discrete regulator of the patterning of intercellular coupling in the zebrafish cardiac epithelium. Thus, we show that systematic prioritization in zebrafish can accelerate the pace of developmental gene function discovery.

Prediction of protein-RNA residue-base contacts using two-dimensional conditional random field with the lasso

Background

To uncover molecular functions and networks in biological cellular systems, it is important to dissect interactions between proteins and RNAs. Many studies have been performed to investigate and analyze interactions between protein amino acid residues and RNA bases. In terms of interactions between residues in proteins, it is generally accepted that an amino acid residue at interacting sites has coevolved together with the partner residue in order to keep the interaction between residues in proteins. Based on this hypothesis, in our previous study to identify residue-residue contact pairs in interacting proteins, we made calculations of mutual information (M I) between amino acid residues from some multiple sequence alignment of homologous proteins, and combined it with a discriminative random field (DRF) approach, which is a special type of conditional random fields (CRFs) and has been proved useful for the purpose of extracting distinguishing areas from a photograph in the image processing field. Recently, the evolutionary correlation of interactions between residues and DNA bases has also been found in certain transcription factors and the DNA-binding sites.

Results

In this paper, we employ more generic two-dimensional CRFs than such DRFs to predict interactions between protein amino acid residues and RNA bases. In addition, we introduce labels representing kinds of amino acids and bases as local features of a CRF. Furthermore, we examine the utility of L₁-norm regularization (lasso) for the CRF. For evaluation of our method, we use residue-base interactions between several Pfam domains and Rfam entries, conduct cross-validation, and calculate the average AUC (Area under ROC Curve) score. The results suggest that our CRF-based method using mutual information and labels with the lasso is useful for further improving the performance, especially provided that the features of CRF are successfully reduced by the lasso approach.

Conclusions

We propose simple and generic two-dimensional CRF models using labels and mutual information with the lasso. Use of the CRF-based method in combination with the lasso is particularly useful for predicting the residue-base contacts in protein-RNA interactions.

Hidden information revealed by optimal community structure from a protein-complex bipartite network improves protein function prediction

The task of extracting the maximal amount of information from a biological network has drawn much attention from researchers, for example, predicting the function of a protein from a protein-protein interaction (PPI) network. It is well known that biological networks consist of modules/communities, a set of nodes that are more densely inter-connected among themselves than with the rest of the network. However, practical applications of utilizing the community information have been rather limited. For protein function prediction on a network, it has been shown that none of the existing community-based protein function prediction methods outperform a simple neighbor-based method. Recently, we have shown that proper utilization of a highly optimal modularity community structure for protein function prediction can outperform neighbor-assisted methods. In this study, we propose two function prediction approaches on bipartite networks that consider the community structure information as well as the neighbor information from the network: 1) a simple screening method and 2) a random forest based method. We demonstrate that our community-assisted methods outperform neighbor-assisted methods and the random forest method yields the best performance. In addition, we show that using the optimal community structure information is essential for more accurate function prediction for the protein-complex bipartite network of Saccharomyces cerevisiae. Community detection can be carried out either using a modified modularity for dealing with the original bipartite network or first projecting the network into a single-mode network (i.e., PPI network) and then applying community detection to the reduced network. We find that the projection leads to the loss of information in a significant way. Since our prediction methods rely only on the network topology, they can be applied to various fields where an efficient network-based analysis is required.

Active subnetwork recovery with a mechanism-dependent scoring function; with application to angiogenesis and organogenesis studies

Background

The learning active subnetworks problem involves finding subnetworks of a bio-molecular network that are active in a particular condition. Many approaches integrate observation data (e.g., gene expression) with the network topology to find candidate subnetworks. Increasingly, pathway databases contain additional annotation information that can be mined to improve prediction accuracy, e.g., interaction mechanism (e.g., transcription, microRNA, cleavage) annotations. We introduce a mechanism-based approach to active subnetwork recovery which exploits such annotations. We suggest that neighboring interactions in a network tend to be co-activated in a way that depends on the “correlation” of their mechanism annotations. e.g., neighboring phosphorylation and de-phosphorylation interactions may be more likely to be co-activated than neighboring phosphorylation and covalent bonding interactions.

Results

Our method iteratively learns the mechanism correlations and finds the most likely active subnetwork. We use a probabilistic graphical model with a Markov Random Field component which creates dependencies between the states (active or non-active) of neighboring interactions, that incorporates a mechanism-based component to the function. We apply a heuristic-based EM-based algorithm suitable for the problem. We validated our method’s performance using simulated data in networks downloaded from GeneGO against the same approach without the mechanism-based component, and two other existing methods. We validated our methods performance in correctly recovering (1) the true interaction states, and (2) global network properties of the original network against these other methods. We applied our method to networks generated from time-course gene expression studies in angiogenesis and lung organogenesis and validated the findings from a biological perspective against current literature.

Conclusions

The advantage of our mechanism-based approach is best seen in networks composed of connected regions with a large number of interactions annotated with a subset of mechanisms, e.g., a regulatory region of transcription interactions, or a cleavage cascade region. When applied to real datasets, our method recovered novel and biologically meaningful putative interactions, e.g., interactions from an integrin signaling pathway using the angiogenesis dataset, and a group of regulatory microRNA interactions in an organogenesis network.

A fast ranking algorithm for predicting gene functions in biomolecular networks

Ranking genes in functional networks according to a specific biological function is a challenging task raising relevant performance and computational complexity problems. To cope with both these problems we developed a transductive gene ranking method based on kernelized score functions able to fully exploit the topology and the graph structure of biomolecular networks and to capture significant functional relationships between genes. We run the method on a network constructed by integrating multiple biomolecular data sources in the yeast model organism, achieving significantly better results than the compared state-of-the-art network-based algorithms for gene function prediction, and with relevant savings in computational time. The proposed approach is general and fast enough to be in perspective applied to other relevant node ranking problems in large and complex biological networks.

Predicting Protein Functions from Protein Interaction Networks

Functional characterization of genes and their protein products is essential to biological and clinical research. Yet, there is still no reliable way of assigning functional annotations to proteins in a high-throughput manner. In this article, the authors provide an introduction to the task of automated protein function prediction. They discuss about the motivation for automated protein function prediction, the challenges faced in this task, as well as some approaches that are currently available. In particular, they take a closer look at methods that use protein-protein interaction for protein function prediction, elaborating on their underlying techniques and assumptions, as well as their strengths and limitations.

A Resource of Quantitative Functional Annotation for Homo sapiens Genes

The body of human genomic and proteomic evidence continues to grow at ever-increasing rates, while annotation efforts struggle to keep pace. A surprisingly small fraction of human genes have clear, documented associations with specific functions, and new functions continue to be found for characterized genes. Here we assembled an integrated collection of diverse genomic and proteomic data for 21,341 human genes and make quantitative associations of each to 4333 Gene Ontology terms. We combined guilt-by-profiling and guilt-by-association approaches to exploit features unique to the data types. Performance was evaluated by cross-validation, prospective validation, and by manual evaluation with the biological literature. Functional-linkage networks were also constructed, and their utility was demonstrated by identifying candidate genes related to a glioma FLN using a seed network from genome-wide association studies. Our annotations are presented—alongside existing validated annotations—in a publicly accessible and searchable web interface.

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.

Network approaches to the functional analysis of microbial proteins

Large amounts of detailed biological data have been generated over the past few decades. Much of these data is freely available in over 1000 online databases; an enticing, but frustrating resource for microbiologists interested in a systems-level view of the structure and function of microbial cells. The frustration engendered by the need to trawl manually through hundreds of databases in order to accumulate information about a gene, protein, pathway, or organism of interest can be alleviated by the use of computational data integration to generated network views of the system of interest. Biological networks can be constructed from a single type of data, such as protein-protein binding information, or from data generated by multiple experimental approaches. In an integrated network, nodes usually represent genes or gene products, while edges represent some form of interaction between the nodes. Edges between nodes may be weighted to represent the probability that the edge exists in vivo. Networks may also be enriched with ontological annotations, facilitating both visual browsing and computational analysis via web service interfaces. In this review, we describe the construction, analysis of both single-data source and integrated networks, and their application to the inference of protein function in microbes.

FUNCTION PREDICTION OF UNCHARACTERIZED PROTEINS

Function prediction of uncharacterized protein sequences generated by genome projects has emerged as an important focus for computational biology. We have categorized several approaches beyond traditional sequence similarity that utilize the overwhelmingly large amounts of available data for computational function prediction, including structure-, association (genomic context)-, interaction (cellular context)-, process (metabolic context)-, and proteomics-experiment-based methods. Because they incorporate structural and experimental data that is not used in sequence-based methods, they can provide additional accuracy and reliability to protein function prediction. Here, first we review the definition of protein function. Then the recent developments of these methods are introduced with special focus on the type of predictions that can be made. The need for further development of comprehensive systems biology techniques that can utilize the ever-increasing data presented by the genomics and proteomics communities is emphasized. For the readers' convenience, tables of useful online resources in each category are included. The role of computational scientists in the near future of biological research and the interplay between computational and experimental biology are also addressed.

Using the underlying biological organization of the Mycobacterium tuberculosis functional network for protein function prediction

Despite ever-increasing amounts of sequence and functional genomics data, there is still a deficiency of functional annotation for many newly sequenced proteins. For Mycobacterium tuberculosis (MTB), more than half of its genome is still uncharacterized, which hampers the search for new drug targets within the bacterial pathogen and limits our understanding of its pathogenicity. As for many other genomes, the annotations of proteins in the MTB proteome were generally inferred from sequence homology, which is effective but its applicability has limitations. We have carried out large-scale biological data integration to produce an MTB protein functional interaction network. Protein functional relationships were extracted from the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database, and additional functional interactions from microarray, sequence and protein signature data. The confidence level of protein relationships in the additional functional interaction data was evaluated using a dynamic data-driven scoring system. This functional network has been used to predict functions of uncharacterized proteins using Gene Ontology (GO) terms, and the semantic similarity between these terms measured using a state-of-the-art GO similarity metric. To achieve better trade-off between improvement of quality, genomic coverage and scalability, this prediction is done by observing the key principles driving the biological organization of the functional network. This study yields a new functionally characterized MTB strain CDC1551 proteome, consisting of 3804 and 3698 proteins out of 4195 with annotations in terms of the biological process and molecular function ontologies, respectively. These data can contribute to research into the Development of effective anti-tubercular drugs with novel biological mechanisms of action.

An Integrated Probabilistic Model for Assessing a Nanocomponent's Reliability

We construct an integrated probabilistic model to capture interactions between atoms of a nanocomponent. We then use this model to assess reliabilities of nanocomponents with different structures. Several properties of our proposed model are also described under a sparseness condition. The model is an extension of our previous model based on Markovian random field theory. The proposed integrated model is flexible in that pairwise relationship information among atoms as well as features of individual atoms can be easily incorporated. An important feature that distinguishes the integrated probabilistic model from our previous model is that the integrated approach uses all available sources of information with different weights for different types of interaction. In this paper we consider the nanocomponent at a fixed moment of time, say the present moment, and we assume that the present state of the nanocomponent depends only on the present states of its atoms.

Conditional random field approach to prediction of protein-protein interactions using domain information

Background

For understanding cellular systems and biological networks, it is important to analyze functions and interactions of proteins and domains. Many methods for predicting protein-protein interactions have been developed. It is known that mutual information between residues at interacting sites can be higher than that at non-interacting sites. It is based on the thought that amino acid residues at interacting sites have coevolved with those at the corresponding residues in the partner proteins. Several studies have shown that such mutual information is useful for identifying contact residues in interacting proteins.

Results

We propose novel methods using conditional random fields for predicting protein-protein interactions. We focus on the mutual information between residues, and combine it with conditional random fields. In the methods, protein-protein interactions are modeled using domain-domain interactions. We perform computational experiments using protein-protein interaction datasets for several organisms, and calculate AUC (Area Under ROC Curve) score. The results suggest that our proposed methods with and without mutual information outperform EM (Expectation Maximization) method proposed by Deng et al., which is one of the best predictors based on domain-domain interactions.

Conclusions

We propose novel methods using conditional random fields with and without mutual information between domains. Our methods based on domain-domain interactions are useful for predicting protein-protein interactions.

Protein annotation from protein interaction networks and Gene Ontology

We introduce a novel method for annotating protein function that combines Naïve Bayes and association rules, and takes advantage of the underlying topology in protein interaction networks and the structure of graphs in the Gene Ontology. We apply our method to proteins from the Human Protein Reference Database (HPRD) and show that, in comparison with other approaches, it predicts protein functions with significantly higher recall with no loss of precision. Specifically, it achieves 51% precision and 60% recall versus 45% and 26% for Majority and 24% and 61% for χ²-statistics, respectively.

Scoring protein relationships in functional interaction networks predicted from sequence data

The abundance of diverse biological data from various sources constitutes a rich source of knowledge, which has the power to advance our understanding of organisms. This requires computational methods in order to integrate and exploit these data effectively and elucidate local and genome wide functional connections between protein pairs, thus enabling functional inferences for uncharacterized proteins. These biological data are primarily in the form of sequences, which determine functions, although functional properties of a protein can often be predicted from just the domains it contains. Thus, protein sequences and domains can be used to predict protein pair-wise functional relationships, and thus contribute to the function prediction process of uncharacterized proteins in order to ensure that knowledge is gained from sequencing efforts. In this work, we introduce information-theoretic based approaches to score protein-protein functional interaction pairs predicted from protein sequence similarity and conserved protein signature matches. The proposed schemes are effective for data-driven scoring of connections between protein pairs. We applied these schemes to the Mycobacterium tuberculosis proteome to produce a homology-based functional network of the organism with a high confidence and coverage. We use the network for predicting functions of uncharacterised proteins.

AVAILABILITY

Protein pair-wise functional relationship scores for Mycobacterium tuberculosis strain CDC1551 sequence data and python scripts to compute these scores are available at http://web.cbio.uct.ac.za/~gmazandu/scoringschemes.

Protein function prediction: towards integration of similarity metrics

Genomic centers discover increasingly many protein sequences and structures, but not necessarily their full biological functions. Thus, currently, less than one percent of proteins have experimentally verified biochemical activities. To fill this gap, function prediction algorithms apply metrics of similarity between proteins on the premise that those sufficiently alike in sequence, or structure, will perform identical functions. Although high sensitivity is elusive, network analyses that integrate these metrics together hold the promise of rapid gains in function prediction specificity.

Predicting node characteristics from molecular networks

A large number of genome-scale networks, including protein-protein and genetic interaction networks, are now available for several organisms. In parallel, many studies have focused on analyzing, characterizing, and modeling these networks. Beyond investigating the topological characteristics such as degree distribution, clustering coefficient, and average shortest-path distance, another area of particular interest is the prediction of nodes (genes) with a given characteristic (labels) - for example prediction of genes that cause a particular phenotype or have a given function. In this chapter, we describe methods and algorithms for predicting node labels from network-based datasets with an emphasis on label propagation algorithms (LPAs) and their relation to local neighborhood methods.

Genome-wide computational function prediction of Arabidopsis proteins by integration of multiple data sources

Although Arabidopsis (Arabidopsis thaliana) is the best studied plant species, the biological role of one-third of its proteins is still unknown. We developed a probabilistic protein function prediction method that integrates information from sequences, protein-protein interactions, and gene expression. The method was applied to proteins from Arabidopsis. Evaluation of prediction performance showed that our method has improved performance compared with single source-based prediction approaches and two existing integration approaches. An innovative feature of our method is that it enables transfer of functional information between proteins that are not directly associated with each other. We provide novel function predictions for 5,807 proteins. Recent experimental studies confirmed several of the predictions. We highlight these in detail for proteins predicted to be involved in flowering and floral organ development.

Network-based auto-probit modeling for protein function prediction

Predicting the functional roles of proteins based on various genome-wide data, such as protein-protein association networks, has become a canonical problem in computational biology. Approaching this task as a binary classification problem, we develop a network-based extension of the spatial auto-probit model. In particular, we develop a hierarchical Bayesian probit-based framework for modeling binary network-indexed processes, with a latent multivariate conditional autoregressive Gaussian process. The latter allows for the easy incorporation of protein-protein association network topologies-either binary or weighted-in modeling protein functional similarity. We use this framework to predict protein functions, for functions defined as terms in the Gene Ontology (GO) database, a popular rigorous vocabulary for biological functionality. Furthermore, we show how a natural extension of this framework can be used to model and correct for the high percentage of false negative labels in training data derived from GO, a serious shortcoming endemic to biological databases of this type. Our method performance is evaluated and compared with standard algorithms on weighted yeast protein-protein association networks, extracted from a recently developed integrative database called Search Tool for the Retrieval of INteracting Genes/proteins (STRING). Results show that our basic method is competitive with these other methods, and that the extended method-incorporating the uncertainty in negative labels among the training data-can yield nontrivial improvements in predictive accuracy.