Ionmob: a Python package for prediction of peptide collisional cross-section values.

Area of Science:

• Proteomics
• Analytical Chemistry
• Computational Biology

Background:

• Ion mobility separation (IMS) coupled with mass spectrometry enhances proteomics coverage and throughput.
• Collisional cross-section (CCS) is a reproducible physicochemical property linking ion mobility to peptide characteristics.
• Accurate CCS prediction is crucial for optimizing experimental design and data analysis in proteomics.

Purpose of the Study:

• To develop a generic, data-driven in silico tool for predicting peptide collisional cross-section (CCS) values.
• To create a customizable Python framework (ionmob) for seamless integration into proteomics workflows.
• To expand the scope of CCS prediction to include post-translationally modified peptides and MHC ligands.

Main Methods:

• Developed ionmob, a Python framework for data preparation, training, and CCS prediction.
• Utilized extensive datasets of phosphorylated peptides and MHC ligand sequences for model training.
• Implemented preprocessing routines for both training and inference stages.

Main Results:

• Created ionmob, a versatile Python framework for peptide CCS prediction.
• Expanded CCS prediction capabilities to include phosphorylated peptides and MHC ligands.
• Demonstrated that in silico predicted CCS values enhance confidence in peptide identification through re-scoring.

Conclusions:

• The ionmob framework provides a valuable, adaptable tool for proteomics research.
• In silico predicted CCS values significantly improve the reliability of peptide identification.
• ionmob facilitates tailored experimental design and refined data processing in mass spectrometry-based proteomics.