Gly-PseAAC: Identifying protein lysine glycation through sequences

Area of Science:

• Biochemistry
• Computational Biology
• Proteomics

Background:

• Glycation, a non-enzymatic process, attaches sugars to peptides, distinct from enzymatic glycosylation.
• Glycation competes with glycosylation, impacting protein function and occurring both in vitro and in vivo.
• Predicting glycated residues is challenging due to a lack of discernible patterns in sequence data.

Purpose of the Study:

• To develop a more effective computational method for predicting glycation sites in peptides.
• To investigate the utility of sequence order information and Position Specific Amino Acid Propensity (PSAAP) for glycation prediction.

Main Methods:

• Exploration of sequence order information and PSAAP for glycation residue identification.
• Development of a Support Vector Machine (SVM) model using PSAAP features.
• Validation of the model's performance using metrics like overall accuracy, Area under the ROC curve, and Matthew's Correlation Coefficient.

Main Results:

• PSAAP effectively discriminates glycated residues from background peptides.
• The SVM model achieved 68.91% overall accuracy.
• The model obtained an Area under the ROC of 0.7258 and a Matthew's Correlation Coefficient of 0.3198.

Conclusions:

• The PSAAP feature set provides a valuable approach for classifying glycation residues.
• The developed computational method offers a practical tool for glycation site prediction.