Residue Adjacency Matrix Based Feature Engineering for Predicting Cysteine Reactivity in Proteins.

ABSTRACT: Free radicals that form from reactive species of nitrogen and oxygen can react dangerously with cellular components and are involved with the pathogenesis of diabetes, cancer, Parkinson's, and heart disease. Cysteine amino acids, due to their reactive nature, are prone to oxidation by these free radicals. Determining which cysteines oxidize within proteins is crucial to our understanding of these chronic diseases. Wet lab techniques, like differential alkylation, to determine which cysteines oxidize are often expensive and time-consuming. We utilize machine learning as a fast and inexpensive approach to identifying cysteines with oxidative capabilities. We created the original features RAMmod and RAMseq for use in classification. We also incorporated well-known features such as PROPKA, SAS

SUBMITTER: Mapes NJ 

PROVIDER: S-EPMC6327741 | biostudies-literature | 2019

REPOSITORIES: biostudies-literature