Prediction of Lysine Glycation PTM site in Protein using Peptide Sequence Evolution based Features

Glycation is a post-translational modification which is non-enzymatic in nature. It is closely associated with different biological functions and responsible for many diseases, for example, diabetes, renal failure etc. Identification of Glycation sites is very important in the development of drugs and the research areas but identifying manually in laboratory is laborious, costly and time consuming. Development of a computational tool will be very useful for the Glycation sites prediction with high accuracy. In our experiment, a new feature extraction technique, called peptide sequence evolution based feature representation, is introduced which gave an Accuracy of 95.94 ± 0.54%, a Sensitivity of 98.20% and a Specificity of 90.67 ± t.07% after running to-fold cross-validation five times. This result outperforms the previously developed tools BPB_GlySite, NetGlycate, PreGly and Gly_ PseAAC.