Loading…

Epik: pKa and Protonation State Prediction through Machine Learning

Epik version 7 is a software program that uses machine learning for predicting the pKa values and protonation state distribution of complex, druglike molecules. Using an ensemble of atomic graph convolutional neural networks (GCNNs) trained on over 42,000 pKa values across broad chemical space from...

Full description

Saved in:
Bibliographic Details
Published in:Journal of chemical theory and computation 2023-04, Vol.19 (8), p.2380-2388
Main Authors: Johnston, Ryne C, Yao, Kun, Kaplan, Zachary, Chelliah, Monica, Leswing, Karl, Seekins, Sean, Watts, Shawn, Calkins, David, Elk, Jackson Chief, Jerome, Steven V, Repasky, Matthew P, Shelley, John C
Format: Article
Language:English
Subjects:
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Epik version 7 is a software program that uses machine learning for predicting the pKa values and protonation state distribution of complex, druglike molecules. Using an ensemble of atomic graph convolutional neural networks (GCNNs) trained on over 42,000 pKa values across broad chemical space from both experimental and computed origins, the model predicts pKa values with 0.42 and 0.72 pKa unit median absolute and root mean square errors, respectively, across seven test sets. Epik version 7 also generates protonation states and recovers 95% of the most populated protonation states compared to previous versions. Requiring on average only 47 ms per ligand, Epik version 7 is rapid and accurate enough to evaluate protonation states for crucial molecules and prepare ultra-large libraries of compounds to explore vast regions of chemical space. The simplicity and time required for the training allow for the generation of highly accurate models customized to a program's specific chemistry.
ISSN:1549-9618
1549-9626
DOI:10.1021/acs.jctc.3c00044