CLASSIFYING PROTEIN S-FARNESYLATION SITES WITH SUPPORT VECTOR MACHINE AND DECISION TREE

Abstract

Protein prenylation is the addition of hydrophobic molecules to a protein or a chemical compound. It is a post-translational modification that plays very important roles affecting to many cellular processes as well as many other cellular functions. Protein S-farnesyl cysteine prenylation is a specific kind of prenylation related to the transfer of a farnesyl moiety to a  cytoplasmic cysteine at or near the C-terminus of the target protein. Recent findings have exhibited the very important roles of S-Farnesyl Cysteine Prenylation (SFCP) that affect to many biological processes as well as have involed in many current common diseases. So far, there has been some researches on SFCP, and several computational tools have been proposed for the classification, prediction of SFCP sites. However, almost of them have not met our demand on related extensive knowlegde, or the predictive performance has not met the requirements. Therefore, in this work, we are motivated to propose an approach to classify protein SFCP based on the incorporation of support vector machine and decision tree. Various features have been investigated to generate the optimal model that has highest predictive performance. The obtained results have demonstrated its ability and feasiblity in the classification of SFCP sites. This could be a suggestion on an approach that can useful for researchers regarding to SFCP.