TopoICSim: a new semantic similarity measure based on gene ontology

Abstract

BackgroundThe Gene Ontology (GO) is a dynamic, controlled vocabulary that describes the cellular function of genes and proteins according to tree major categories: biological process, molecular function and cellular component. It has become widely used in many bioinformatics applications for annotating genes and measuring their semantic similarity, rather than their sequence similarity. Generally speaking, semantic similarity measures involve the GO tree topology, information content of GO terms, or a combination of both.ResultsHere we present a new semantic similarity measure called TopoICSim (Topological Information Content Similarity) which uses information on the specific paths between GO terms based on the topology of the GO tree, and the distribution of information content along these paths. The TopoICSim algorithm was evaluated on two human benchmark datasets based on KEGG pathways and Pfam domains grouped as clans, using GO terms from either the biological process or molecular function. The performance of the TopoICSim measure compared favorably to five existing methods. Furthermore, the TopoICSim similarity was also tested on gene/protein sets defined by correlated gene expression, using three human datasets, and showed improved performance compared to two previously published similarity measures. Finally we used an online benchmarking resource which evaluates any similarity measure against a set of 11 similarity measures in three tests, using gene/protein sets based on sequence similarity, Pfam domains, and enzyme classifications. The results for TopoICSim showed improved performance relative to most of the measures included in the benchmarking, and in particular a very robust performance throughout the different tests.ConclusionsThe TopoICSim similarity measure provides a competitive method with robust performance for quantification of semantic similarity between genes and proteins based on GO annotations. An R script for TopoICSim is available at http://bigr.medisin.ntnu.no/tools/TopoICSim.R.

Towards a semantic approach to numerical tree inference in phylogenetics

“…Related applications exist for measuring semantic similarity and are used for the mapping of correspondences and the matching of similarities between concepts of different ontologies (see, e.g. Noy et al., ; Zhang and Bodenreider, ,b; Parmentier et al., ; Travillian et al., ; Bertone et al., ; see also Hoehndorf et al., ; for instance matching see also Shao et al., ; see also Topological Information Content Similarity , Ehsani and Drabløs, ; but see also ontology‐driven approach for analysing molecular networks, Landeghem et al., ). The result of this initial step is the identification of units of comparison (= comparative homologues , Vogt, ) across the semantic instance anatomies of all sister taxa for a given possible tree, which is documented in a consensus comparison graph .…”

“…Ontology-based annotations of morphological 2D and 3D images could be utilized as another data source (e.g. Dadzie and Burger, 2005;Larson and Martone, 2009;Eliceiri et al, 2012), which could not only significantly contribute to the size of the available data, but also contribute valuable image references for documentation, which subsequently can be used for visualizing the results of the numerical tree inference step (cf. Kuß et al, 2008).…”

Metabolic Pathway Membership Inference using an Ontology-based Similarity Approach

“…The third approach combine the node-and edge-based approaches. Essentially, this hybrid approach considers both the information contained in the nodes and the structure of the GO graph using path distance as weight (Chen et al (2007), Wu et al (2013), Ehsani & Drabløs (2016)), in order to minimize the problem found in edgeand nod-based approach alone.…”

“…TopoICSim (Ehsani & Drabløs (2016)) is a hybrid-based similiraty measure which uses both IC and topology of GO terms in measuring term similarity. First it calculates the weighted distance between two terms.…”

Combining Semantic Similarity and GO Enrichment for Computation of Functional Similarity

“…Semantic similarity measures the similarity of two ontology terms by typically evaluating their commonness normalized to their uniqueness in terms of information contents (Harispe et al 2014) (Ranwez et al 2014). The commonness of two terms is typically evaluated by the information content of the lowest/closest common ancestor as used by Resnik (Resnik 1999), Lin (Lin 1998), Nunivers (Mazandu & Mulder 2013), relevance similarity (Schlicker et al 2006) measures; or by the information content of all common ancestors as evaluated by XGraSM (Couto & Silva 2011) and TopoICSim (Ehsani & Drabløs 2016). The uniqueness of GO terms are often evaluated by taking the average of the information content (IC) of the two terms.…”