{"database":"biostudies-literature","file_versions":[],"scores":null,"additional":{"omics_type":["Unknown"],"submitter":["Tariq U"],"funding":["NIGMS NIH HHS"],"pubmed_abstract":["Database search algorithms reduce the number of potential candidate peptides against which scoring needs to be performed using a single (i.e. mass) property for filtering. While useful, filtering based on one property may lead to exclusion of non-abundant spectra and uncharacterized peptides - potentially exacerbating the <i>streetlight</i> effect. Here we present <i>ProteoRift</i>, a novel attention and multitask deep-network, which can <i>predict</i> multiple peptide properties (length, missed cleavages, and modification status) directly from spectra. We demonstrate that <i>ProteoRift</i> can predict these properties with up to 97% accuracy resulting in search-space reduction by more than 90%. As a result, our end-to-end pipeline is shown to exhibit 8x to 12x speedups with peptide deduction accuracy comparable to algorithmic techniques. We also formulate two uncertainty estimation metrics, which can distinguish between in-distribution and out-of-distribution data (ROC-AUC 0.99) and predict high-scoring mass spectra against correct peptide (ROC-AUC 0.94). These models and metrics are integrated in an end-to-end ML pipeline available at https://github.com/pcdslab/ProteoRift."],"journal":["bioRxiv : the preprint server for biology"],"pagination":["2024.08.21.609035"],"full_dataset_link":["https://www.ebi.ac.uk/biostudies/studies/S-EPMC11370541"],"repository":["biostudies-literature"],"pubmed_title":["Predicting peptide properties from mass spectrometry data using deep attention-based multitask network and uncertainty quantification."],"pmcid":["PMC11370541"],"funding_grant_id":["R35 GM153434"],"pubmed_authors":["Saeed F","Tariq U"],"additional_accession":[]},"is_claimable":false,"name":"Predicting peptide properties from mass spectrometry data using deep attention-based multitask network and uncertainty quantification.","description":"Database search algorithms reduce the number of potential candidate peptides against which scoring needs to be performed using a single (i.e. mass) property for filtering. While useful, filtering based on one property may lead to exclusion of non-abundant spectra and uncharacterized peptides - potentially exacerbating the <i>streetlight</i> effect. Here we present <i>ProteoRift</i>, a novel attention and multitask deep-network, which can <i>predict</i> multiple peptide properties (length, missed cleavages, and modification status) directly from spectra. We demonstrate that <i>ProteoRift</i> can predict these properties with up to 97% accuracy resulting in search-space reduction by more than 90%. As a result, our end-to-end pipeline is shown to exhibit 8x to 12x speedups with peptide deduction accuracy comparable to algorithmic techniques. We also formulate two uncertainty estimation metrics, which can distinguish between in-distribution and out-of-distribution data (ROC-AUC 0.99) and predict high-scoring mass spectra against correct peptide (ROC-AUC 0.94). These models and metrics are integrated in an end-to-end ML pipeline available at https://github.com/pcdslab/ProteoRift.","dates":{"release":"2024-01-01T00:00:00Z","publication":"2024 Aug","modification":"2026-07-02T03:20:19.768Z","creation":"2025-04-06T10:14:19.231Z"},"accession":"S-EPMC11370541","cross_references":{"pubmed":["39229185"],"doi":["10.1101/2024.08.21.609035"]}}