iPTM-mLys: identifying multiple lysine PTM sites and their different types

Area of Science:

• Bioinformatics
• Computational Biology
• Proteomics

Background:

• Post-translational modifications (PTMs) are crucial for protein function and drug development.
• Existing computational methods predict single-label PTMs but not multi-label PTMs.
• Lysine (K) residues are frequently modified by multiple PTM types, posing a prediction challenge.

Purpose of the Study:

• To develop the first computational predictor for identifying multi-label PTMs on protein sequences.
• To address the challenge of predicting which lysine residues accommodate single, multiple, or no PTMs.

Main Methods:

• Developed a novel multi-label predictor named IPTM-MLYS.
• Incorporated sequence-coupled effects into the general Pseudo Amino Acid Composition (PseAAC).
• Fused multiple basic random forest classifiers into an ensemble system.

Main Results:

• IPTM-MLYS is the first predictor capable of handling multi-label PTM identification.
• Rigorous cross-validation using multi-label metrics demonstrated promising and encouraging performance.
• The predictor successfully addresses the challenge of identifying multiple PTMs on single lysine residues.

Conclusions:

• IPTM-MLYS represents a significant advancement in PTM site prediction.
• The tool facilitates a deeper understanding of protein biological processes.
• A user-friendly web server is available for experimental scientists.