Automatic prediction of lipid metabolism function based on binary SVM classifier in Tetrahymena thermophila

BRACALENTE FERNANDO; COSTA JOAQUIN; ARABOLAZA ANA LORENA; GRAMAJO HUGO; UTTARO ANTONIO; SPETALE FLAVIO EZEQUIEL

Capital Federal

Simposio; 6to Simposio Argentino de Jóvenes Investigadores en Bioinformática; 2021

The ciliate Tetrahymena thermophila, is a simple and experimental malleable organism that makes it a valuable study model for the lipid metabolism in higher eukaryotes. For example, its sterol metabolism is of particular interest to aim to fill the gap in the understanding of sterol import, intracellular transport and signaling systems in different eukaryotic lineages. Prior to the biochemical study of the lipid metabolism in T. thermophila, we intend to reduce the number of putatively involved proteins through in silico analysis but the standard methods for the annotation of protein-coding genes based on sequence similarity, BLASTP, do not give good results. Therefore, automatic functional annotation methods based on machine learning are developed as an alternative. In particular, we will focus on those who use the Gene Ontology (GO) which provides a controlled vocabulary where each term is associated with a functionality.