ionmob

ionmob predicts peptide collisional cross-section (CCS) values to link ion mobility separation (IMS) measurements with mass spectrometry proteomics analyses.

Key Features:

• Data preparation and training: Framework for data preparation, training, and prediction of peptide CCS values.
• Pretrained models and preprocessing routines: Includes pretrained models and preprocessing routines for CCS prediction.
• Extensive peptide database: Incorporates approximately 21,000 unique phosphorylated peptides and around 17,000 MHC ligand sequences with their respective charge state pairs for model training and prediction.

Scientific Applications:

• Experimental design: Predicts peptide CCS values to inform and tailor acquisition methods in mass spectrometry experiments, optimizing coverage and throughput.
• Database search refinement: Uses predicted CCS values to refine database search results and increase the accuracy of peptide identification.
• Re-scoring methods: Enables re-scoring of identified peptides using in silico predicted CCS values to boost confidence in peptide identifications.

Methodology:

Leverages machine learning techniques to predict CCS values from peptide data and is trained on a large dataset of unique peptides, with an emphasis on workflow engineering for applicability and adaptability.