More Agility to Semantic Similarities Algorithm Implementations

Int J Environ Res Public Health. 2019 Dec 30;17(1):267. doi: 10.3390/ijerph17010267.

Abstract

Algorithms for measuring semantic similarity between Gene Ontology (GO) terms has become a popular area of research in bioinformatics as it can help to detect functional associations between genes and potential impact to the health and well-being of humans, animals, and plants. While the focus of the research is on the design and improvement of GO semantic similarity algorithms, there is still a need for implementation of such algorithms before they can be used to solve actual biological problems. This can be challenging given that the potential users usually come from a biology background and they are not programmers. A number of implementations exist for some well-established algorithms but these implementations are not generic enough to support any algorithm other than the ones they are designed for. The aim of this paper is to shift the focus away from implementation, allowing researchers to focus on algorithm's design and execution rather than implementation. This is achieved by an implementation approach capable of understanding and executing user defined GO semantic similarity algorithms. Questions and answers were used for the definition of the user defined algorithm. Additionally, this approach understands any direct acyclic digraph in an Open Biomedical Ontologies (OBO)-like format and its annotations. On the other hand, software developers of similar applications can also benefit by using this as a template for their applications.

Keywords: GO semantic terms similarity; Gene Ontology similarity algorithms; digital health; gene/gene product semantic similarity.

MeSH terms

  • Algorithms*
  • Animals
  • Computational Biology
  • Gene Ontology*
  • Humans
  • Semantics*
  • Software