Output details

Improving GO semantic similarity measures by exploring the ontology beneath the terms and modelling uncertainty

<28>Much work exists discussing similarity measures over ontologies which are DAGs (e.g. Gene Ontology). It was important to analyse their correctness. For the first time we show limitations of existing algorithms and provide an algorithm that can be used to “correct” existing measures and this correction greatly improves the measures. This algorithm was part of the de-noising procedure that we applied to the largest ever human protein interaction network (published in Cell, output 2). We recently developed a web-service implementing this algorithm (under review in Bioinformatics). The algorithm was developed entirely in Paccanaro's lab. Paccanaro is senior corresponding author.