Abstract
Pyrrolidone carboxylic acid (PCA) is formed during a common post-translational modification (PTM) of extracellular and multi-pass membrane proteins. In this study, we developed a new predictor to predict the modification sites of PCA based on maximum relevance minimum redundancy (mRMR) and incremental feature selection (IFS). We incorporated 727 features that belonged to 7 kinds of protein properties to predict the modification sites, including sequence conservation, residual disorder, amino acid factor, secondary structure and solvent accessibility, gain/loss of amino acid during evolution, propensity of amino acid to be conserved at protein-protein interface and protein surface, and deviation of side chain carbon atom number. Among these 727 features, 244 features were selected by mRMR and IFS as the optimized features for the prediction, with which the prediction model achieved a maximum of MCC of 0.7812. Feature analysis showed that all feature types contributed to the modification process. Further site-specific feature analysis showed that the features derived from PCA's surrounding sites contributed more to the determination of PCA sites than other sites. The detailed feature analysis in this paper might provide important clues for understanding the mechanism of the PCA formation and guide relevant experimental validations.
Highlights
Post-translational modifications (PTMs) are crucial for proteins to maintain their structural and functional diversities in both prokaryotes and eukaryotes
A smaller index of a feature suggests that it is more important for the prediction of Pyrrolidone carboxylic acid (PCA) modification site
We investigated the number of each kind of amino acids of the position-specific conversation scoring matrixes (PSSM) features (Figure 3A) and found that the conservation against mutations to the 20 amino acids influents differently on the prediction of PCA modification site
Summary
Post-translational modifications (PTMs) are crucial for proteins to maintain their structural and functional diversities in both prokaryotes and eukaryotes. As a glutamic acid derivative that lacks a H2O molecule [3] in extracellular and multi-pass membrane proteins, many studies have demonstrated that pyroglutamic acid is formed either late in protein translation by cyclization of the glutamine at the N-terminus or as a post-translational event, just prior to the secretion of completed proteins from the cell [5]. Modified proteins of this type usually show an increase half-life, because PCA blocks proteins, minimizing their susceptibility to degradation by aminopeptidases [3,6]. Identification of protein pyroglutamic acid modification sites is of fundamental importance to understand the mechanism by which pGlu occurs in biological systems
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
