DeepLC introduces transfer learning for accurate LC retention time prediction and adaptation to substantially different modifications and setups
Year: 2025
Type: preprint
Abstract: While LC retention time prediction of peptides and their modifications has proven useful, widespread adoption and optimal performance are hindered by variations in experimental parameters. These variations can render retention time prediction models inaccurate and dramatically reduce the value of predictions for identification, validation, and DIA spectral library generation. To date, mitigation of these issues has been attempted through calibration or by training bespoke models for specific experimental setups, with only partial success. We here demonstrate that transfer learning can successfully overcome these limitations by leveraging pre-trained model parameters. Remarkably, this approach can even fit highly performant models to substantially different peptide modifications and LC conditions than those on which the model was originally trained. This impressive adaptability of transfer learning makes it a highly robust solution for accurate peptide retention time prediction across a very wide variety of imaginable proteomics workflows.
Language: English
Subfield: Spectroscopy
Field: Chemistry
Domain: Physical Sciences
Open Access status: green
Funders: Villum Fonden, European Commission, Agence Nationale de la Recherche, Fonds Wetenschappelijk Onderzoek, Lundbeckfonden, Novo Nordisk, Universiteit Gent, Syddansk Universitet, Novo Nordisk Fonden, Vlaamse regering, H. Lundbeck A/S, HORIZON EUROPE Framework Programme, VILLUM Center for Bioanalytical Sciences, CHIST-ERA
Awards: G1024401664, A novel era in proteomics-based protein PTM discovery through deep learning approaches, G1382555053, G2409256426, Putting the AI in DIA - building an ML-driven multi-modification engine, Improved data-driven bioinformatics tools to greatly extend neoand xeno-epitope landscapes detected by immunopeptidomics., Computational Proteomics Training European Innovative Network, G434337789, A diaPASEF powered computational deep dive into the myotonic dystrophy type 1 liquid biopsy proteome, Advanced AI-driven Bioinformatics Solutions for Performant Peptidomics, Immunopeptidomics-based Development of Next-Generation Bacterial mRNA Vaccines, ProteinContour: proteome-scale unraveling of the relation between post-translational modifications, biophysical properties, interactions and sub-cellular location of proteins., G7240651115, Comparative signature of Marburg virus cell activation as a blueprint for the identification of antiviral targets against newly emerging viruses, G8439718446
OpenAlex ID: W4411035750