Tractable and Thermodynamically Consistent pKa Prediction via First-Dissociation Aggregation
- 1School of Pharmacy, Shihezi University
Abstract
Ionization governs molecular behavior, yet predicting aqueous pKa accurately and tractably remains a fundamental challenge. Rigorous ensemble methods require enumerating a protonation-state space that grows exponentially with the number of ionizable sites, while fast graph predictors return numbers without thermodynamic consistency. Here we derive an exact thermodynamic identity showing that the macroscopic first-dissociation constant is fully determined by protonated-microstate populations and unique first-deprotonation events, eliminating the deprotonated ensemble algebraically and reducing aggregation to O(n+E) operations. We implement this identity in DTi-pKa, a dual-head graph neural network whose free-energy and dissociation heads are coupled by thermodynamic constraints. On four external benchmarks (355 molecules), DTi-pKa achieves a pooled MAE of 0.5665 pKa units, with 0.6285 on the 169 records lacking training-cache matches, while providing site-resolved micro-pKa values and protonation-state populations. Controlled ablations show that constraint placement at the level of the reported macroscopic equilibrium matters more than microscopic label precision--a design principle transferable to physics-informed machine learning beyond chemistry. Closure diagnostics expose numerical self-consistency as an open frontier, which we report transparently. By replacing an intractable computation with a physical identity, DTi-pKa unifies macroscopic accuracy and microscopic interpretability in a single framework.
Related
- Time trend prediction and spatial-temporal aggregation analysis of scrub typhus in Ganzhou City, China, during 2008-2021.
- Seq2Pocket: augmenting protein language models for spatially consistent binding site prediction
- An experimentally tractable Arabidopsis-Swede Midge model to explore host reprogramming by gall-inducing insects
- Thermodynamically controlled biokinetics modelling of a lindane-degrading anaerobic enrichment culture
- A rapid, genetically-tractable human iPSC-derived model for studying tissue-resident macrophage biology in Alzheimer's and other diseases
BibTeX
@misc{liu2026tractable,
title = {{Tractable and Thermodynamically Consistent pKa Prediction via First-Dissociation Aggregation}},
author = {Z. Liu and Q. Zhu and Q. Li and Y. Zhu and Y. Xu and B. Dai and Y. Li and Z. Liu and S. Yang},
howpublished = {bioRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.30.755586},
url = {https://www.biorxiv.org/content/10.64898/2026.09.30.755586v1},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - Tractable and Thermodynamically Consistent pKa Prediction via First-Dissociation Aggregation
AU - Z. Liu
AU - Q. Zhu
AU - Q. Li
AU - Y. Zhu
AU - Y. Xu
AU - B. Dai
AU - Y. Li
AU - Z. Liu
AU - S. Yang
PY - 2026
DA - 2026/09/30
PB - bioRxiv
DO - 10.64898/2026.09.30.755586
UR - https://www.biorxiv.org/content/10.64898/2026.09.30.755586v1
AB - Ionization governs molecular behavior, yet predicting aqueous pKa accurately and tractably remains a fundamental challenge. Rigorous ensemble methods require enumerating a protonation-state space that grows exponentially with the number of ionizable sites, while fast graph predictors return numbers without thermodynamic consistency. Here we derive an exact thermodynamic identity showing that the macroscopic first-dissociation constant is fully determined by protonated-microstate populations and unique first-deprotonation events, eliminating the deprotonated ensemble algebraically and reducing aggregation to O(n+E) operations. We implement this identity in DTi-pKa, a dual-head graph neural network whose free-energy and dissociation heads are coupled by thermodynamic constraints. On four external benchmarks (355 molecules), DTi-pKa achieves a pooled MAE of 0.5665 pKa units, with 0.6285 on the 169 records lacking training-cache matches, while providing site-resolved micro-pKa values and protonation-state populations. Controlled ablations show that constraint placement at the level of the reported macroscopic equilibrium matters more than microscopic label precision--a design principle transferable to physics-informed machine learning beyond chemistry. Closure diagnostics expose numerical self-consistency as an open frontier, which we report transparently. By replacing an intractable computation with a physical identity, DTi-pKa unifies macroscopic accuracy and microscopic interpretability in a single framework.
ER -