Parametric Bayesian priors and better choice of negative examples improve protein function prediction

Predicting protein functions using positive-unlabeled ranking with ontology-based priors

Automated protein function prediction is a crucial and widely studied problem in bioinformatics. Computationally, protein function is a multilabel classification problem where only positive samples are defined and there is a large number of unlabeled annotations. Most existing methods rely on the assumption that the unlabeled set of protein function annotations are negatives, inducing the false negative issue, where potential positive samples are trained as negatives. We introduce a novel approach named PU-GO, wherein we address function prediction as a positive-unlabeled ranking problem. We apply empirical risk minimization, i.e. we minimize the classification risk of a classifier where class priors are obtained from the Gene Ontology hierarchical structure. We show that our approach is more robust than other state-of-the-art methods on similarity-based and time-based benchmark datasets.

AVAILABILITY AND IMPLEMENTATION

Data and code are available at https://github.com/bio-ontology-research-group/PU-GO.

Computational Methods for Prediction of Human Protein-Phenotype Associations: A Review

Deciphering the relationship between human proteins (genes) and phenotypes is one of the fundamental tasks in phenomics research. The Human Phenotype Ontology (HPO) builds upon a standardized logical vocabulary to describe the abnormal phenotypes encountered in human diseases and paves the way towards the computational analysis of their genetic causes. To date, many computational methods have been proposed to predict the HPO annotations of proteins. In this paper, we conduct a comprehensive review of the existing approaches to predicting HPO annotations of novel proteins, identifying missing HPO annotations, and prioritizing candidate proteins with respect to a certain HPO term. For each topic, we first give the formalized description of the problem, and then systematically revisit the published literatures highlighting their advantages and disadvantages, followed by the discussion on the challenges and promising future directions. In addition, we point out several potential topics to be worthy of exploration including the selection of negative HPO annotations and detecting HPO misannotations. We believe that this review will provide insight to the researchers in the field of computational phenotype analyses in terms of comprehending and developing novel prediction algorithms.

Accurate and efficient gene function prediction using a multi-bacterial network

MOTIVATION

Nearly 40% of the genes in sequenced genomes have no experimentally or computationally derived functional annotations. To fill this gap, we seek to develop methods for network-based gene function prediction that can integrate heterogeneous data for multiple species with experimentally based functional annotations and systematically transfer them to newly sequenced organisms on a genome-wide scale. However, the large sizes of such networks pose a challenge for the scalability of current methods.

RESULTS

We develop a label propagation algorithm called FastSinkSource. By formally bounding its rate of progress, we decrease the running time by a factor of 100 without sacrificing accuracy. We systematically evaluate many approaches to construct multi-species bacterial networks and apply FastSinkSource and other state-of-the-art methods to these networks. We find that the most accurate and efficient approach is to pre-compute annotation scores for species with experimental annotations, and then to transfer them to other organisms. In this manner, FastSinkSource runs in under 3 min for 200 bacterial species.

AVAILABILITY AND IMPLEMENTATION

An implementation of our framework and all data used in this research are available at https://github.com/Murali-group/multi-species-GOA-prediction.

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.

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.

A Literature Review of Gene Function Prediction by Modeling Gene Ontology

Annotating the functional properties of gene products, i.e., RNAs and proteins, is a fundamental task in biology. The Gene Ontology database (GO) was developed to systematically describe the functional properties of gene products across species, and to facilitate the computational prediction of gene function. As GO is routinely updated, it serves as the gold standard and main knowledge source in functional genomics. Many gene function prediction methods making use of GO have been proposed. But no literature review has summarized these methods and the possibilities for future efforts from the perspective of GO. To bridge this gap, we review the existing methods with an emphasis on recent solutions. First, we introduce the conventions of GO and the widely adopted evaluation metrics for gene function prediction. Next, we summarize current methods of gene function prediction that apply GO in different ways, such as using hierarchical or flat inter-relationships between GO terms, compressing massive GO terms and quantifying semantic similarities. Although many efforts have improved performance by harnessing GO, we conclude that there remain many largely overlooked but important topics for future research.

Improving protein function prediction using protein sequence and GO-term similarities

MOTIVATION

Most automatic functional annotation methods assign Gene Ontology (GO) terms to proteins based on annotations of highly similar proteins. We advocate that proteins that are less similar are still informative. Also, despite their simplicity and structure, GO terms seem to be hard for computers to learn, in particular the Biological Process ontology, which has the most terms (>29 000). We propose to use Label-Space Dimensionality Reduction (LSDR) techniques to exploit the redundancy of GO terms and transform them into a more compact latent representation that is easier to predict.

RESULTS

We compare proteins using a sequence similarity profile (SSP) to a set of annotated training proteins. We introduce two new LSDR methods, one based on the structure of the GO, and one based on semantic similarity of terms. We show that these LSDR methods, as well as three existing ones, improve the Critical Assessment of Functional Annotation performance of several function prediction algorithms. Cross-validation experiments on Arabidopsis thaliana proteins pinpoint the superiority of our GO-aware LSDR over generic LSDR. Our experiments on A.thaliana proteins show that the SSP representation in combination with a kNN classifier outperforms state-of-the-art and baseline methods in terms of cross-validated F-measure.

AVAILABILITY AND IMPLEMENTATION

Source code for the experiments is available at https://github.com/stamakro/SSP-LSDR.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

deepNF: deep network fusion for protein function prediction

Motivation

The prevalence of high-throughput experimental methods has resulted in an abundance of large-scale molecular and functional interaction networks. The connectivity of these networks provides a rich source of information for inferring functional annotations for genes and proteins. An important challenge has been to develop methods for combining these heterogeneous networks to extract useful protein feature representations for function prediction. Most of the existing approaches for network integration use shallow models that encounter difficulty in capturing complex and highly non-linear network structures. Thus, we propose deepNF, a network fusion method based on Multimodal Deep Autoencoders to extract high-level features of proteins from multiple heterogeneous interaction networks.

Results

We apply this method to combine STRING networks to construct a common low-dimensional representation containing high-level protein features. We use separate layers for different network types in the early stages of the multimodal autoencoder, later connecting all the layers into a single bottleneck layer from which we extract features to predict protein function. We compare the cross-validation and temporal holdout predictive performance of our method with state-of-the-art methods, including the recently proposed method Mashup. Our results show that our method outperforms previous methods for both human and yeast STRING networks. We also show substantial improvement in the performance of our method in predicting gene ontology terms of varying type and specificity.

Availability and implementation

deepNF is freely available at: https://github.com/VGligorijevic/deepNF.

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.

Evaluating the impact of topological protein features on the negative examples selection

BACKGROUND

Supervised machine learning methods when applied to the problem of automated protein-function prediction (AFP) require the availability of both positive examples (i.e., proteins which are known to possess a given protein function) and negative examples (corresponding to proteins not associated with that function). Unfortunately, publicly available proteome and genome data sources such as the Gene Ontology rarely store the functions not possessed by a protein. Thus the negative selection, consisting in identifying informative negative examples, is currently a central and challenging problem in AFP. Several heuristics have been proposed through the years to solve this problem; nevertheless, despite their effectiveness, to the best of our knowledge no previous existing work studied which protein features are more relevant to this task, that is, which protein features help more in discriminating reliable and unreliable negatives.

RESULTS

The present work analyses the impact of several features on the selection of negative proteins for the Gene Ontology (GO) terms. The analysis is network-based: it exploits the fact that proteins can be naturally structured in a network, considering the pairwise relationships coming from several sources of data, such as protein-protein and genetic interactions. Overall, the proposed protein features, including local and global graph centrality measures and protein multifunctionality, can be term-aware (i.e., depending on the GO term) and term-unaware (i.e., invariant across the GO terms). We validated the informativeness of each feature utilizing a temporal holdout in three different experiments on yeast, mouse and human proteomes: (i) feature selection to detect which protein features are more helpful for the negative selection; (ii) protein function prediction to verify whether the features considered are also useful to predict GO terms; (iii) negative selection by applying two different negative selection algorithms on proteins represented through the proposed features.

CONCLUSIONS

Term-aware features (with some exceptions) resulted more informative for problem (i), together with node betweenness, which is the most relevant among term-unaware features. The node positive neighborhood instead is the most predictive feature for the AFP problem, while experiment (iii) showed that the proposed features allow negative selection algorithms to select effectively negative instances in the temporal holdout setting, with better results when nonlinear combinations of features are also exploited.

BUSCA: an integrative web server to predict subcellular localization of proteins

Here, we present BUSCA (http://busca.biocomp.unibo.it), a novel web server that integrates different computational tools for predicting protein subcellular localization. BUSCA combines methods for identifying signal and transit peptides (DeepSig and TPpred3), GPI-anchors (PredGPI) and transmembrane domains (ENSEMBLE3.0 and BetAware) with tools for discriminating subcellular localization of both globular and membrane proteins (BaCelLo, MemLoci and SChloro). Outcomes from the different tools are processed and integrated for annotating subcellular localization of both eukaryotic and bacterial protein sequences. We benchmark BUSCA against protein targets derived from recent CAFA experiments and other specific data sets, reporting performance at the state-of-the-art. BUSCA scores better than all other evaluated methods on 2732 targets from CAFA2, with a F1 value equal to 0.49 and among the best methods when predicting targets from CAFA3. We propose BUSCA as an integrated and accurate resource for the annotation of protein subcellular localization.

Large-scale automated function prediction of protein sequences and an experimental case study validation on PTEN transcript variants

Recent advances in computing power and machine learning empower functional annotation of protein sequences and their transcript variations. Here, we present an automated prediction system UniGOPred, for GO annotations and a database of GO term predictions for proteomes of several organisms in UniProt Knowledgebase (UniProtKB). UniGOPred provides function predictions for 514 molecular function (MF), 2909 biological process (BP), and 438 cellular component (CC) GO terms for each protein sequence. UniGOPred covers nearly the whole functionality spectrum in Gene Ontology system and it can predict both generic and specific GO terms. UniGOPred was run on CAFA2 challenge target protein sequences and it is categorized within the top 10 best performing methods for the molecular function category. In addition, the performance of UniGOPred is higher compared to the baseline BLAST classifier in all categories of GO. UniGOPred predictions are compared with UniProtKB/TrEMBL database annotations as well. Furthermore, the proposed tool's ability to predict negatively associated GO terms that defines the functions that a protein does not possess, is discussed. UniGOPred annotations were also validated by case studies on PTEN protein variants experimentally and on CHD8 protein variants with literature. UniGOPred protein functional annotation system is available as an open access tool at http://cansyl.metu.edu.tr/UniGOPred.html.

NoGOA: predicting noisy GO annotations using evidences and sparse representation

BACKGROUND

Gene Ontology (GO) is a community effort to represent functional features of gene products. GO annotations (GOA) provide functional associations between GO terms and gene products. Due to resources limitation, only a small portion of annotations are manually checked by curators, and the others are electronically inferred. Although quality control techniques have been applied to ensure the quality of annotations, the community consistently report that there are still considerable noisy (or incorrect) annotations. Given the wide application of annotations, however, how to identify noisy annotations is an important but yet seldom studied open problem.

RESULTS

We introduce a novel approach called NoGOA to predict noisy annotations. NoGOA applies sparse representation on the gene-term association matrix to reduce the impact of noisy annotations, and takes advantage of sparse representation coefficients to measure the semantic similarity between genes. Secondly, it preliminarily predicts noisy annotations of a gene based on aggregated votes from semantic neighborhood genes of that gene. Next, NoGOA estimates the ratio of noisy annotations for each evidence code based on direct annotations in GOA files archived on different periods, and then weights entries of the association matrix via estimated ratios and propagates weights to ancestors of direct annotations using GO hierarchy. Finally, it integrates evidence-weighted association matrix and aggregated votes to predict noisy annotations. Experiments on archived GOA files of six model species (H. sapiens, A. thaliana, S. cerevisiae, G. gallus, B. Taurus and M. musculus) demonstrate that NoGOA achieves significantly better results than other related methods and removing noisy annotations improves the performance of gene function prediction.

CONCLUSIONS

The comparative study justifies the effectiveness of integrating evidence codes with sparse representation for predicting noisy GO annotations. Codes and datasets are available at http://mlda.swu.edu.cn/codes.php?name=NoGOA .

NegGOA: negative GO annotations selection using ontology structure

MOTIVATION

Predicting the biological functions of proteins is one of the key challenges in the post-genomic era. Computational models have demonstrated the utility of applying machine learning methods to predict protein function. Most prediction methods explicitly require a set of negative examples-proteins that are known not carrying out a particular function. However, Gene Ontology (GO) almost always only provides the knowledge that proteins carry out a particular function, and functional annotations of proteins are incomplete. GO structurally organizes more than tens of thousands GO terms and a protein is annotated with several (or dozens) of these terms. For these reasons, the negative examples of a protein can greatly help distinguishing true positive examples of the protein from such a large candidate GO space.

RESULTS

In this paper, we present a novel approach (called NegGOA) to select negative examples. Specifically, NegGOA takes advantage of the ontology structure, available annotations and potentiality of additional annotations of a protein to choose negative examples of the protein. We compare NegGOA with other negative examples selection algorithms and find that NegGOA produces much fewer false negatives than them. We incorporate the selected negative examples into an efficient function prediction model to predict the functions of proteins in Yeast, Human, Mouse and Fly. NegGOA also demonstrates improved accuracy than these comparing algorithms across various evaluation metrics. In addition, NegGOA is less suffered from incomplete annotations of proteins than these comparing methods.

AVAILABILITY AND IMPLEMENTATION

The Matlab and R codes are available at https://sites.google.com/site/guoxian85/neggoa

CONTACT

gxyu@swu.edu.cn

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Negative example selection for protein function prediction: the NoGO database

Negative examples – genes that are known not to carry out a given protein function – are rarely recorded in genome and proteome annotation databases, such as the Gene Ontology database. Negative examples are required, however, for several of the most powerful machine learning methods for integrative protein function prediction. Most protein function prediction efforts have relied on a variety of heuristics for the choice of negative examples. Determining the accuracy of methods for negative example prediction is itself a non-trivial task, given that the Open World Assumption as applied to gene annotations rules out many traditional validation metrics. We present a rigorous comparison of these heuristics, utilizing a temporal holdout, and a novel evaluation strategy for negative examples. We add to this comparison several algorithms adapted from Positive-Unlabeled learning scenarios in text-classification, which are the current state of the art methods for generating negative examples in low-density annotation contexts. Lastly, we present two novel algorithms of our own construction, one based on empirical conditional probability, and the other using topic modeling applied to genes and annotations. We demonstrate that our algorithms achieve significantly fewer incorrect negative example predictions than the current state of the art, using multiple benchmarks covering multiple organisms. Our methods may be applied to generate negative examples for any type of method that deals with protein function, and to this end we provide a database of negative examples in several well-studied organisms, for general use (The NoGO database, available at: bonneaulab.bio.nyu.edu/nogo.html).

Many machine learning methods have been applied to the task of predicting the biological function of proteins based on a variety of available data. The majority of these methods require negative examples: proteins that are known not to perform a function, in order to achieve meaningful predictions, but negative examples are often not available. In addition, past heuristic methods for negative example selection suffer from a high error rate. Here, we rigorously compare two novel algorithms against past heuristics, as well as some algorithms adapted from a similar task in text-classification. Through this comparison, performed on several different benchmarks, we demonstrate that our algorithms make significantly fewer mistakes when predicting negative examples. We also provide a database of negative examples for general use in machine learning for protein function prediction (The NoGO database, available at: bonneaulab.bio.nyu.edu/nogo.html).

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.

Network analysis: a new approach to study endocrine disorders

Systems biology is the study of the interactions that occur between the components of individual cells - including genes, proteins, transcription factors, small molecules, and metabolites, and their relationships to complex physiological and pathological processes. The application of systems biology to medicine promises rapid advances in both our understanding of disease and the development of novel treatment options. Network biology has emerged as the primary tool for studying systems biology as it utilises the mathematical analysis of the relationships between connected objects in a biological system and allows the integration of varied 'omic' datasets (including genomics, metabolomics, proteomics, etc.). Analysis of network biology generates interactome models to infer and assess function; to understand mechanisms, and to prioritise candidates for further investigation. This review provides an overview of network methods used to support this research and an insight into current applications of network analysis applied to endocrinology. A wide spectrum of endocrine disorders are included ranging from congenital hyperinsulinism in infancy, through childhood developmental and growth disorders, to the development of metabolic diseases in early and late adulthood, such as obesity and obesity-related pathologies. In addition to providing a deeper understanding of diseases processes, network biology is also central to the development of personalised treatment strategies which will integrate pharmacogenomics with systems biology of the individual.

High-resolution functional annotation of human transcriptome: predicting isoform functions by a novel multiple instance-based label propagation method

Alternative transcript processing is an important mechanism for generating functional diversity in genes. However, little is known about the precise functions of individual isoforms. In fact, proteins (translated from transcript isoforms), not genes, are the function carriers. By integrating multiple human RNA-seq data sets, we carried out the first systematic prediction of isoform functions, enabling high-resolution functional annotation of human transcriptome. Unlike gene function prediction, isoform function prediction faces a unique challenge: the lack of the training data--all known functional annotations are at the gene level. To address this challenge, we modelled the gene-isoform relationships as multiple instance data and developed a novel label propagation method to predict functions. Our method achieved an average area under the receiver operating characteristic curve of 0.67 and assigned functions to 15 572 isoforms. Interestingly, we observed that different functions have different sensitivities to alternative isoform processing, and that the function diversity of isoforms from the same gene is positively correlated with their tissue expression diversity. Finally, we surveyed the literature to validate our predictions for a number of apoptotic genes. Strikingly, for the famous 'TP53' gene, we not only accurately identified the apoptosis regulation function of its five isoforms, but also correctly predicted the precise direction of the regulation.

Progress and challenges in the computational prediction of gene function using networks: 2012-2013 update

In an opinion published in 2012, we reviewed and discussed our studies of how gene network-based guilt-by-association (GBA) is impacted by confounds related to gene multifunctionality. We found such confounds account for a significant part of the GBA signal, and as a result meaningfully evaluating and applying computationally-guided GBA is more challenging than generally appreciated. We proposed that effort currently spent on incrementally improving algorithms would be better spent in identifying the features of data that do yield novel functional insights. We also suggested that part of the problem is the reliance by computational biologists on gold standard annotations such as the Gene Ontology. In the year since, there has been continued heavy activity in GBA-based research, including work that contributes to our understanding of the issues we raised. Here we provide a review of some of the most relevant recent work, or which point to new areas of progress and challenges.