{"database":"biostudies-literature","file_versions":[],"scores":null,"additional":{"omics_type":["Unknown"],"submitter":["Shen M"],"funding":["NCI NIH HHS","NIGMS NIH HHS"],"pubmed_abstract":["Despite its importance in understanding biology and computer-aided drug discovery, the accurate prediction of protein ionization states remains a formidable challenge. Physics-based approaches struggle to capture the small, competing contributions in the complex protein environment, while machine learning (ML) is hampered by scarcity of experimental data. Here we report the development of p<i>K</i> <sub>a</sub> ML (KaML) models based on decision trees and graph attention networks (GAT), exploiting physicochemical understanding and a new experiment p<i>K</i> <sub>a</sub> database (PKAD-3) enriched with highly shifted p<i>K</i> <sub>a</sub>'s. KaML-CBtree significantly outperforms the current state of the art in predicting p<i>K</i> <sub>a</sub> values and ionization states across all six ti"],"journal":["bioRxiv : the preprint server for biology"],"pagination":["2024.11.09.622800"],"full_dataset_link":["https://www.ebi.ac.uk/biostudies/studies/S-EPMC11601431"],"repository":["biostudies-literature"],"pubmed_title":["KaMLs for Predicting Protein p<i>K</i> <sub>a</sub> Values and Ionization States: Are Trees All You Need?"],"pmcid":["PMC11601431"],"funding_grant_id":["R01 CA256557","R35 GM148261"],"pubmed_authors":["Shen M","Bhatnagar S","Shen J","Ambrozak S","Liu R","Kortzak D","Buchanan I"],"additional_accession":[]},"is_claimable":false,"name":"KaMLs for Predicting Protein p<i>K</i> <sub>a</sub> Values and Ionization States: Are Trees All You Need?","description":"Despite its importance in understanding biology and computer-aided drug discovery, the accurate prediction of protein ionization states remains a formidable challenge. Physics-based approaches struggle to capture the small, competing contributions in the complex protein environment, while machine learning (ML) is hampered by scarcity of experimental data. Here we report the development of p<i>K</i> <sub>a</sub> ML (KaML) models based on decision trees and graph attention networks (GAT), exploiting physicochemical understanding and a new experiment p<i>K</i> <sub>a</sub> database (PKAD-3) enriched with highly shifted p<i>K</i> <sub>a</sub>'s. KaML-CBtree significantly outperforms the current state of the art in predicting p<i>K</i> <sub>a</sub> values and ionization states across all six ti","dates":{"release":"2025-01-01T00:00:00Z","publication":"2025 Jan","modification":"2026-04-19T03:22:59.488Z","creation":"2025-04-06T01:56:37.723Z"},"accession":"S-EPMC11601431","cross_references":{"pubmed":["39605739"],"doi":["10.1101/2024.11.09.622800"]}}