GOssTo & GOssToWeb: user-friendly tools for calculating semantic similarities on the Gene Ontology

Semantic similarity measures have become important in bioinformatics as they quantify relatedness between genes in a way that is complementary to both experimental information and sequence-based approaches. Several semantic similarity measures have been proposed for the Gene Ontology. Recently, Yang et.al.(2012) introduced the Random Walk Contribution, a kind of “add-on” for any existing semantic similarity measure: a measure to take into account the structure of the GO DAG beneath the terms as well as the uncertainty of GO annotations. Yang et. al. were able to show a large increase in the performance of existing semantic similarity measures when they were combined with their Random Walk Contribution. In this poster we present GOssTO (Gene Ontology semantic similarity Tool) and its online counterpart GOssTOWeb, two new tools for calculating semantic similarities. GOssTo and GOssTOWeb support both graph-based as well as term-based similarity measures and implement Yang’s et.al.(2012) Random Walk Contribution.