UNIPred: Unbalance-Aware Network Integration and Prediction of Protein Functions

Frasca, Marco; Bertoni, Alberto; Valentini, Giorgio

doi:10.1089/cmb.2014.0110

Cited by 18 publications

(21 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Then we selected the GO terms with at least 20 proteins in C np , obtaining the terms summarized in Table 2. The proposed features were also tested in terms of their capability in predicting the protein functions, by selecting GO terms with 20–200 annotations in the later release, in order to have a minimum of information to train a classifier, and to exclude terms with a large number of annotations, because they are too generic [13, 24, 25]. The total number of obtained GO terms is shown in Table 3.…”

Section: Methodsmentioning

confidence: 99%

“…That is, before applying any algorithm to learn negative examples, it is of paramount importance studying which ‘protein representation’ is more informative for the problem itself. In this context, most information sources about the relationships between proteins are naturally represented through protein networks, where each node represents a protein and an edge the relationship between two proteins [13]; additionally, most approaches proposed for AFP are network-based [14–20]. Thus, the purpose here is twofold: extracting meaningful protein features from protein networks, and assessing their ability to improve the identification of good negative examples.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

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

2018

Self Cite

View full text Add to dashboard Cite

BackgroundSupervised 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.ResultsThe 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.ConclusionsTerm-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.Electronic supplementary materialThe online version of this article (10.1186/s12859-018-2385-x) contains supplementary material, which is available to authorized users.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

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

2018

Self Cite

View full text Add to dashboard Cite

show abstract

“…Although COSNET has already been validated in [14, 15, 37] to solve AFP, for sake of completeness we report in Table 2 its performances in predicting the GO terms. The generalization abilities of COSNET have been assessed through a 5-fold cross validation (CV), and evaluated in terms of Precision (the proportion of positives correctly predicted) and Recall (the proportion of real positive discovered) combined in the F measure, which is the harmonic mean of precision and recall.…”

Section: Resultsmentioning

confidence: 99%

A GPU-based algorithm for fast node label learning in large and unbalanced biomolecular networks

et al. 2018

Self Cite

View full text Add to dashboard Cite

BackgroundSeveral problems in network biology and medicine can be cast into a framework where entities are represented through partially labeled networks, and the aim is inferring the labels (usually binary) of the unlabeled part. Connections represent functional or genetic similarity between entities, while the labellings often are highly unbalanced, that is one class is largely under-represented: for instance in the automated protein function prediction (AFP) for most Gene Ontology terms only few proteins are annotated, or in the disease-gene prioritization problem only few genes are actually known to be involved in the etiology of a given disease. Imbalance-aware approaches to accurately predict node labels in biological networks are thereby required. Furthermore, such methods must be scalable, since input data can be large-sized as, for instance, in the context of multi-species protein networks.ResultsWe propose a novel semi-supervised parallel enhancement of COSNet, an imbalance-aware algorithm build on Hopfield neural model recently suggested to solve the AFP problem. By adopting an efficient representation of the graph and assuming a sparse network topology, we empirically show that it can be efficiently applied to networks with millions of nodes. The key strategy to speed up the computations is to partition nodes into independent sets so as to process each set in parallel by exploiting the power of GPU accelerators. This parallel technique ensures the convergence to asymptotically stable attractors, while preserving the asynchronous dynamics of the original model. Detailed experiments on real data and artificial big instances of the problem highlight scalability and efficiency of the proposed method.ConclusionsBy parallelizing COSNet we achieved on average a speed-up of 180x in solving the AFP problem in the S. cerevisiae, Mus musculus and Homo sapiens organisms, while lowering memory requirements. In addition, to show the potential applicability of the method to huge biomolecular networks, we predicted node labels in artificially generated sparse networks involving hundreds of thousands to millions of nodes.

show abstract

“…Finally, recent studies have improved the GP performance by integrating multiple data sources, including expression profiles, SNP genotype data, expression quantitative trait loci, and so on [19]: network-based approaches indeed construct a consensus network which includes the specificity of each network, covers more genes and contains more accurate pairwise connections [20].…”

Section: Firstmentioning

confidence: 99%

Exploiting Negative Sample Selection for Prioritizing Candidate Disease Genes

Frasca

Malchiodi

2017

Genomics Comput Biol

Self Cite

View full text Add to dashboard Cite

A major challenge in bio-medicine is finding the genetic causes of human diseases, and researchers are often faced with a large number of candidate genes. Gene prioritization methods provide a valuable support in guiding researchers to detect reliable candidate causative-genes for a disease under study. Indeed, such methods rank genes according to their association with a disease of interest. Actually, the majority of genetic disorders has few or none causative genes associated with them; this induces a high labeling unbalance in the corresponding ranking problems, thus linking the need of achieving reliable solutions to the adoption of imbalance-aware techniques. We propose the use of an expressly designed imbalance-aware methodology for prioritizing genes, which first rebalances the training set entries through a negative selection procedure, then applies a learning algorithm 'sensitive' to the misclassification of positive instances, to provide the gene ranking. The algorithm has a reduced time complexity, which makes feasible its application on large-sized datasets. The validation of this methodology proved its competitiveness with state-of-art techniques on a benchmark composed of 708 selected Medical Subject Headings diseases, and provided some putative novel gene-disease associations.

show abstract

UNIPred: Unbalance-Aware Network Integration and Prediction of Protein Functions

Cited by 18 publications

References 56 publications

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

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

A GPU-based algorithm for fast node label learning in large and unbalanced biomolecular networks

Exploiting Negative Sample Selection for Prioritizing Candidate Disease Genes

Contact Info

Product

Resources

About