Improving Small Molecule pKa Prediction Using Transfer Learning With Graph Neural Networks
Authors:
Open Access
- 26 May 2022
- journal article
- research article
- Published by Frontiers Media SA in Frontiers in Chemistry
- Vol. 10, 866585
Abstract
Enumerating protonation states and calculating microstate pKa values of small molecules is an important yet challenging task for lead optimization and molecular modeling. Commercial and non-commercial solutions have notable limitations such as restrictive and expensive licenses, high CPU/GPU hour requirements, or the need for expert knowledge to set up and use. We present a graph neural network model that is trained on 714,906 calculated microstate pKa predictions from molecules obtained from the ChEMBL database. The model is fine-tuned on a set of 5,994 experimental pKa values significantly improving its performance on two challenging test sets. Combining the graph neural network model with Dimorphite-DL, an open-source program for enumerating ionization states, we have developed the open-source Python package pkasolver, which is able to generate and enumerate protonation states and calculate pKa values with high accuracy.Keywords
All Related Versions
Funding Information
- Austrian Science Fund (J 4245-N28)
- Innovative Medicines Initiative (IMI2 JU 821528)
- Horizon 2020 Framework Programme
This publication has 35 references indexed in Scilit:
- Substituent effects on the basicity (pK a ) of aryl guanidines and 2-(arylimino)imidazolidines: correlations of pH-metric and UV-metric values with predictions from gas-phase ab initio bond lengthsNew Journal of Chemistry, 2017
- ChEMBL web services: streamlining access to drug discovery data and utilitiesNucleic Acids Research, 2015
- The significance of acid/base properties in drug discoveryChemical Society Reviews, 2012
- ChEMBL: a large-scale bioactivity database for drug discoveryNucleic Acids Research, 2011
- Predicting the pKa of Small MoleculesCombinatorial Chemistry & High Throughput Screening, 2011
- Extended-Connectivity FingerprintsJournal of Chemical Information and Modeling, 2010
- Towards the comprehensive, rapid, and accurate prediction of the favorable tautomeric states of drug-like molecules in aqueous solutionJournal of Computer-Aided Molecular Design, 2010
- Comparison of Nine Programs Predicting pKa Values of Pharmaceutical SubstancesJournal of Chemical Information and Modeling, 2009
- Epik: a software program for pK a prediction and protonation state generation for drug-like moleculesJournal of Computer-Aided Molecular Design, 2007
- pK a Prediction for Organic Acids and Bases1981