Incorporating functional inter-relationships into protein function prediction algorithms

Background  

Functional classification schemes (e.g. the Gene Ontology) that serve as the basis for annotation efforts in several organisms are often the source of gold standard information for computational efforts at supervised protein function prediction. While successful function prediction algorithms have been developed, few previous efforts have utilized more than the protein-to-functional class label information provided by such knowledge bases. For instance, the Gene Ontology not only captures protein annotations to a set of functional classes, but it also arranges these classes in a DAG-based hierarchy that captures rich inter-relationships between different classes. These inter-relationships present both opportunities, such as the potential for additional training examples for small classes from larger related classes, and challenges, such as a harder to learn distinction between similar GO terms, for standard classification-based approaches.     

Functional annotation prediction: all for one and one for all

In an era of rapid genome sequencing and high-throughput technology, automatic function prediction for a novel sequence is of utter importance in bioinformatics. While automatic annotation methods based on local alignment searches can be simple and straightforward, they suffer from several drawbacks, including relatively low sensitivity and assignment of incorrect annotations that are not associated with the region of similarity. ProtoNet is a hierarchical organization of the protein sequences in the UniProt database. Although the hierarchy is constructed in an unsupervised automatic manner, it has been shown to be coherent with several biological data sources. We extend the ProtoNet system in order to assign functional annotations automatically. By leveraging on the scaffold of the hierarchical classification, the method is able to overcome some frequent annotation pitfalls.     

Predicting gene function using few positive examples and unlabeled ones

Background

A large amount of functional genomic data have provided enough knowledge in predicting gene function computationally, which uses known functional annotations and relationship between unknown genes and known ones to map unknown genes to GO functional terms. The prediction procedure is usually formulated as binary classification problem. Training binary classifier needs both positive examples and negative ones that have almost the same size. However, from various annotation database, we can only obtain few positive genes annotation for most offunctional terms, that is, there are only few positive examples for training classifier, which makes predicting directly gene function infeasible.

Results

We propose a novel approach SPE_RNE to train classifier for each functional term. Firstly, positive examples set is enlarged by creating synthetic positive examples. Secondly, representative negative examples are selected by training SVM(support vector machine) iteratively to move classification hyperplane to a appropriate place. Lastly, an optimal SVM classifier are trained by using grid search technique. On combined kernel ofYeast protein sequence, microarray expression, protein-protein interaction and GO functional annotation data, we compare SPE_RNE with other three typical methods in three classical performance measures recall R, precise P and their combination F: twoclass considers all unlabeled genes as negative examples, twoclassbal selects randomly same number negative examples from unlabeled gene, PSoL selects a negative examples set that are far from positive examples and far from each other.

Conclusions

In test data and unknown genes data, we compute average and variant of measure F. The experiments showthat our approach has better generalized performance and practical prediction capacity. In addition, our method can also be used for other organisms such as human.

     

Functional Knowledge Transfer for High-accuracy Prediction of Under-studied Biological Processes

A key challenge in genetics is identifying the functional roles of genes in pathways. Numerous functional genomics techniques (e.g. machine learning) that predict protein function have been developed to address this question. These methods generally build from existing annotations of genes to pathways and thus are often unable to identify additional genes participating in processes that are not already well studied. Many of these processes are well studied in some organism, but not necessarily in an investigator''s organism of interest. Sequence-based search methods (e.g. BLAST) have been used to transfer such annotation information between organisms. We demonstrate that functional genomics can complement traditional sequence similarity to improve the transfer of gene annotations between organisms. Our method transfers annotations only when functionally appropriate as determined by genomic data and can be used with any prediction algorithm to combine transferred gene function knowledge with organism-specific high-throughput data to enable accurate function prediction.We show that diverse state-of-art machine learning algorithms leveraging functional knowledge transfer (FKT) dramatically improve their accuracy in predicting gene-pathway membership, particularly for processes with little experimental knowledge in an organism. We also show that our method compares favorably to annotation transfer by sequence similarity. Next, we deploy FKT with state-of-the-art SVM classifier to predict novel genes to 11,000 biological processes across six diverse organisms and expand the coverage of accurate function predictions to processes that are often ignored because of a dearth of annotated genes in an organism. Finally, we perform in vivo experimental investigation in Danio rerio and confirm the regulatory role of our top predicted novel gene, wnt5b, in leftward cell migration during heart development. FKT is immediately applicable to many bioinformatics techniques and will help biologists systematically integrate prior knowledge from diverse systems to direct targeted experiments in their organism of study.     

CrowdGO: Machine learning and semantic similarity guided consensus Gene Ontology annotation

Characterising gene function for the ever-increasing number and diversity of species with annotated genomes relies almost entirely on computational prediction methods. These software are also numerous and diverse, each with different strengths and weaknesses as revealed through community benchmarking efforts. Meta-predictors that assess consensus and conflict from individual algorithms should deliver enhanced functional annotations. To exploit the benefits of meta-approaches, we developed CrowdGO, an open-source consensus-based Gene Ontology (GO) term meta-predictor that employs machine learning models with GO term semantic similarities and information contents. By re-evaluating each gene-term annotation, a consensus dataset is produced with high-scoring confident annotations and low-scoring rejected annotations. Applying CrowdGO to results from a deep learning-based, a sequence similarity-based, and two protein domain-based methods, delivers consensus annotations with improved precision and recall. Furthermore, using standard evaluation measures CrowdGO performance matches that of the community’s best performing individual methods. CrowdGO therefore offers a model-informed approach to leverage strengths of individual predictors and produce comprehensive and accurate gene functional annotations.     

A framework for incorporating functional interrelationships into protein function prediction algorithms

The functional annotation of proteins is one of the most important tasks in the post-genomic era. Although many computational approaches have been developed in recent years to predict protein function, most of these traditional algorithms do not take interrelationships among functional terms into account, such as different GO terms usually coannotate with some common proteins. In this study, we propose a new functional similarity measure in the form of Jaccard coefficient to quantify these interrelationships and also develop a framework for incorporating GO term similarity into protein function prediction process. The experimental results of cross-validation on S. cerevisiae and Homo sapiens data sets demonstrate that our method is able to improve the performance of protein function prediction. In addition, we find that small size terms associated with a few of proteins obtain more benefit than the large size ones when considering functional interrelationships. We also compare our similarity measure with other two widely used measures, and results indicate that when incorporated into function prediction algorithms, our proposed measure is more effective. Experiment results also illustrate that our algorithms outperform two previous competing algorithms, which also take functional interrelationships into account, in prediction accuracy. Finally, we show that our method is robust to annotations in the database which are not complete at present. These results give new insights about the importance of functional interrelationships in protein function prediction.     

A categorization approach to automated ontological function annotation

Automated function prediction (AFP) methods increasingly use knowledge discovery algorithms to map sequence, structure, literature, and/or pathway information about proteins whose functions are unknown into functional ontologies, typically (a portion of) the Gene Ontology (GO). While there are a growing number of methods within this paradigm, the general problem of assessing the accuracy of such prediction algorithms has not been seriously addressed. We present first an application for function prediction from protein sequences using the POSet Ontology Categorizer (POSOC) to produce new annotations by analyzing collections of GO nodes derived from annotations of protein BLAST neighborhoods. We then also present hierarchical precision and hierarchical recall as new evaluation metrics for assessing the accuracy of any predictions in hierarchical ontologies, and discuss results on a test set of protein sequences. We show that our method provides substantially improved hierarchical precision (measure of predictions made that are correct) when applied to the nearest BLAST neighbors of target proteins, as compared with simply imputing that neighborhood's annotations to the target. Moreover, when our method is applied to a broader BLAST neighborhood, hierarchical precision is enhanced even further. In all cases, such increased hierarchical precision performance is purchased at a modest expense of hierarchical recall (measure of all annotations that get predicted at all).     

Automatic annotation of protein function

The annotation of protein function at genomic scale is essential for day-to-day work in biology and for any systematic approach to the modeling of biological systems. Currently, functional annotation is essentially based on the expansion of the relatively small number of experimentally determined functions to large collections of proteins. The task of systematic annotation faces formidable practical problems related to the accuracy of the input experimental information, the reliability of current systems for transferring information between related sequences, and the reproducibility of the links between database information and the original experiments reported in publications. These technical difficulties merely lie on the surface of the deeper problem of the evolution of protein function in the context of protein sequences and structures. Given the mixture of technical and scientific challenges, it is not surprising that errors are introduced, and expanded, in database annotations. In this situation, a more realistic option is the development of a reliability index for database annotations, instead of depending exclusively on efforts to correct databases. Several groups have attempted to compare the database annotations of similar proteins, which constitutes the first steps toward the calibration of the relationship between sequence and annotation space.     

Functional annotation from predicted protein interaction networks

MOTIVATION: Progress in large-scale experimental determination of protein-protein interaction networks for several organisms has resulted in innovative methods of functional inference based on network connectivity. However, the amount of effort and resources required for the elucidation of experimental protein interaction networks is prohibitive. Previously we, and others, have developed techniques to predict protein interactions for novel genomes using computational methods and data generated from other genomes. RESULTS: We evaluated the performance of a network-based functional annotation method that makes use of our predicted protein interaction networks. We show that this approach performs equally well on experimentally derived and predicted interaction networks, for both manually and computationally assigned annotations. We applied the method to predicted protein interaction networks for over 50 organisms from all domains of life, providing annotations for many previously unannotated proteins and verifying existing low-confidence annotations. AVAILABILITY: Functional predictions for over 50 organisms are available at http://bioverse.compbio.washington.edu and datasets used for analysis at http://data.compbio.washington.edu/misc/downloads/nannotation_data/. SUPPLEMENTARY INFORMATION: A supplemental appendix gives additional details not in the main text. (http://data.compbio.washington.edu/misc/downloads/nannotation_data/supplement.pdf).     

Quality of computationally inferred gene ontology annotations

Gene Ontology (GO) has established itself as the undisputed standard for protein function annotation. Most annotations are inferred electronically, i.e. without individual curator supervision, but they are widely considered unreliable. At the same time, we crucially depend on those automated annotations, as most newly sequenced genomes are non-model organisms. Here, we introduce a methodology to systematically and quantitatively evaluate electronic annotations. By exploiting changes in successive releases of the UniProt Gene Ontology Annotation database, we assessed the quality of electronic annotations in terms of specificity, reliability, and coverage. Overall, we not only found that electronic annotations have significantly improved in recent years, but also that their reliability now rivals that of annotations inferred by curators when they use evidence other than experiments from primary literature. This work provides the means to identify the subset of electronic annotations that can be relied upon-an important outcome given that >98% of all annotations are inferred without direct curation.     

Automated genome sequence analysis and annotation.

MOTIVATION: Large-scale genome projects generate a rapidly increasing number of sequences, most of them biochemically uncharacterized. Research in bioinformatics contributes to the development of methods for the computational characterization of these sequences. However, the installation and application of these methods require experience and are time consuming. RESULTS: We present here an automatic system for preliminary functional annotation of protein sequences that has been applied to the analysis of sets of sequences from complete genomes, both to refine overall performance and to make new discoveries comparable to those made by human experts. The GeneQuiz system includes a Web-based browser that allows examination of the evidence leading to an automatic annotation and offers additional information, views of the results, and links to biological databases that complement the automatic analysis. System structure and operating principles concerning the use of multiple sequence databases, underlying sequence analysis tools, lexical analyses of database annotations and decision criteria for functional assignments are detailed. The system makes automatic quality assessments of results based on prior experience with the underlying sequence analysis tools; overall error rates in functional assignment are estimated at 2.5-5% for cases annotated with highest reliability ('clear' cases). Sources of over-interpretation of results are discussed with proposals for improvement. A conservative definition for reporting 'new findings' that takes account of database maturity is presented along with examples of possible kinds of discoveries (new function, family and superfamily) made by the system. System performance in relation to sequence database coverage, database dynamics and database search methods is analysed, demonstrating the inherent advantages of an integrated automatic approach using multiple databases and search methods applied in an objective and repeatable manner. AVAILABILITY: The GeneQuiz system is publicly available for analysis of protein sequences through a Web server at http://www.sander.ebi.ac. uk/gqsrv/submit     

Functional Inferences from Blind ab Initio Protein Structure Predictions

Ab initio protein structure prediction methods have improved dramatically in the past several years. Because these methods require only the sequence of the protein of interest, they are potentially applicable to the open reading frames in the many organisms whose sequences have been and will be determined. Ab initio methods cannot currently produce models of high enough resolution for use in rational drug design, but there is an exciting potential for using the methods for functional annotation of protein sequences on a genomic scale. Here we illustrate how functional insights can be obtained from low-resolution predicted structures using examples from blind ab initio structure predictions from the third and fourth critical assessment of structure prediction (CASP3, CASP4) experiments.     

Sources of systematic error in functional annotation of genomes: domain rearrangement, non-orthologous gene displacement and operon disruption

Functional annotation of proteins encoded in newly sequenced genomes can be expected to meet two conflicting objectives: (i) provide as much information as possible, and (ii) avoid erroneous functional assignments and over-predictions. The continuing exponential growth of the number of sequenced genomes makes the quality of sequence annotation a critical factor in the efforts to utilize this new information. When dubious functional assignments are used as a basis for subsequent predictions, they tend to proliferate, leading to "database explosion". It is therefore important to identify the common factors that hamper functional annotation. As a first step towards that goal, we have compared the annotations of the Mycoplasma genitalium and Methanococcus jannaschii genomes produced in several independent studies. The most common causes of questionable predictions appear to be: i) non-critical use of annotations from existing database entries; ii) taking into account only the annotation of the best database hit; iii) insufficient masking of low complexity regions (e.g. non-globular domains) in protein sequences, resulting in spurious database hits obscuring relevant ones; iv) ignoring multi-domain organization of the query proteins and/or the database hits; v) non-critical functional inferences on the basis of the functions of neighboring genes in an operon; vi) non-orthologous gene displacement, i.e. involvement of structurally unrelated proteins in the same function. These observations suggest that case by case validation of functional annotation by expert biologists remains crucial for productive genome analysis.     

COLLAPSE: A representation learning framework for identification and characterization of protein structural sites

The identification and characterization of the structural sites which contribute to protein function are crucial for understanding biological mechanisms, evaluating disease risk, and developing targeted therapies. However, the quantity of known protein structures is rapidly outpacing our ability to functionally annotate them. Existing methods for function prediction either do not operate on local sites, suffer from high false positive or false negative rates, or require large site‐specific training datasets, necessitating the development of new computational methods for annotating functional sites at scale. We present COLLAPSE (Compressed Latents Learned from Aligned Protein Structural Environments), a framework for learning deep representations of protein sites. COLLAPSE operates directly on the 3D positions of atoms surrounding a site and uses evolutionary relationships between homologous proteins as a self‐supervision signal, enabling learned embeddings to implicitly capture structure–function relationships within each site. Our representations generalize across disparate tasks in a transfer learning context, achieving state‐of‐the‐art performance on standardized benchmarks (protein–protein interactions and mutation stability) and on the prediction of functional sites from the prosite database. We use COLLAPSE to search for similar sites across large protein datasets and to annotate proteins based on a database of known functional sites. These methods demonstrate that COLLAPSE is computationally efficient, tunable, and interpretable, providing a general‐purpose platform for computational protein analysis.     

Functional inferences from blind ab initio protein structure predictions

Ab initio protein structure prediction methods have improved dramatically in the past several years. Because these methods require only the sequence of the protein of interest, they are potentially applicable to the open reading frames in the many organisms whose sequences have been and will be determined. Ab initio methods cannot currently produce models of high enough resolution for use in rational drug design, but there is an exciting potential for using the methods for functional annotation of protein sequences on a genomic scale. Here we illustrate how functional insights can be obtained from low-resolution predicted structures using examples from blind ab initio structure predictions from the third and fourth critical assessment of structure prediction (CASP3, CASP4) experiments.     

Proteogenomics: needs and roles to be filled by proteomics in genome annotation

While genome sequencing efforts reveal the basic building blocksof life, a genome sequence alone is insufficient for elucidatingbiological function. Genome annotation—the process ofidentifying genes and assigning function to each gene in a genomesequence—provides the means to elucidate biological functionfrom sequence. Current state-of-the-art high-throughput genomeannotation uses a combination of comparative (sequence similaritydata) and non-comparative (ab initio gene prediction algorithms)methods to identify protein-coding genes in genome sequences.Because approaches used to validate the presence of predictedprotein-coding genes are typically based on expressed RNA sequences,they cannot independently and unequivocally determine whethera predicted protein-coding gene is translated into a protein.With the ability to directly measure peptides arising from expressedproteins, high-throughput liquid chromatography-tandem massspectrometry-based proteomics approaches can be used to verifycoding regions of a genomic sequence. Here, we highlight severalways in which high-throughput tandem mass spectrometry-basedproteomics can improve the quality of genome annotations andsuggest that it could be efficiently applied during the genecalling process so that the improvements are propagated throughthe subsequent functional annotation process.