Computational modeling and optimized machine learning for multidrug solubility prediction in supercritical carbon dioxide
- 1Center of Research Excellence in Renewable Energy and Power Systems/Energy Efficiency Group, Department of Mechanical Engineering, Faculty of Engineering, King Abdulaziz University , ,
- 2Mechanical Engineering Department, College of Engineering and Architecture, Umm Al-Qura University , ,
- 3Department of Mechanical Engineering, Faculty of Engineering, King Abdulaziz University , ,
Abstract
Reliable prediction of pharmaceutical solubility in supercritical carbon dioxide (scCO2) is complicated by strong compound-specific behavior and response values. This study investigated whether sparse measurements from a newly introduced drug can calibrate transferable machine-learning predictions for computational estimation of solubility in wide range of parameters. A curated workbook containing 1,618 observations for 60 drugs was modeled using molecular weight, melting point, temperature, and pressure. Solubility was transformed as log₁₀(S), and XGBoost, CatBoost, and multilayer perceptron regression were each tuned using tree-structured Parzen estimation, differential evolution, and particle swarm optimization. Evaluation used five-fold nested group cross-validation, with complete drugs held out from training and tuning. For each test drug, predictions were assessed before calibration and after 1, 3, or 5 randomly selected measurements, repeated ten times. Within the present evaluation framework, the lowest-log-MAE zero-shot configuration was XGBoost optimized by tree-structured Parzen estimation (log-MAE = 0.584; R² = −0.087; RMSE = 1.713 g/L), showing that the four descriptors were insufficient for reliable extrapolation to unseen compounds. Sparse calibration substantially improved transfer: the lowest-log-MAE one-shot result was obtained by CatBoost–particle swarm optimization (log-MAE = 0.418; R² = 0.166), whereas XGBoost–tree-structured Parzen estimation had the lowest log-MAE at three shots (log-MAE = 0.298; R² = 0.317) and at five shots (log-MAE = 0.251; R² = 0.452; RMSE = 1.223 g/L). Median absolute relative deviation decreased from 82.4% to 42.9% between zero and five shots. These results establish sparse, leakage-controlled calibration as a more defensible use case than unconditional prediction for unseen drugs, while identifying the need for richer molecular descriptors and external prospective validation. In practical terms, this few-shot strategy can reduce the experimental burden for a newly introduced drug by using a small screening set to calibrate a global model before mapping the remaining pressure-temperature domain.
Related
- AI-driven predictive modeling of supercritical CO₂ solubility using cheetah-optimized machine learning.
- Physics-consistent equation discovery for drug solubility in supercritical CO₂ through symbolic regression and domain generalization
- Symbolic and domain-generalized machine learning for interpretable solubility modeling in supercritical CO₂
- HSPiP, computational, and thermodynamic models-based solubility validation of tofacitinib in PEG400 – water system, shelf-life estimation, and GastroPlus-enabled predictive in vivo performance
- Optimized alkyne-modified pyridylpiperazine efflux pump inhibitors overcome intrinsic macrolide resistance in multidrug-resistant Klebsiella pneumoniae isolates
BibTeX
@article{abuhamdeh2026computational,
title = {{Computational modeling and optimized machine learning for multidrug solubility prediction in supercritical carbon dioxide}},
author = {Nidal H. Abu-Hamdeh and Hasan H. Hijji and Khalid H. Almitani},
journal = {Frontiers in medicine},
year = {2026},
month = oct,
doi = {10.3389/fmed.2026.1985801},
url = {https://europepmc.org/article/PMC/PMC13719417},
note = {CC BY},
}
RIS
TY - JOUR
TI - Computational modeling and optimized machine learning for multidrug solubility prediction in supercritical carbon dioxide
AU - Nidal H. Abu-Hamdeh
AU - Hasan H. Hijji
AU - Khalid H. Almitani
PY - 2026
DA - 2026/10/16
JO - Frontiers in medicine
DO - 10.3389/fmed.2026.1985801
UR - https://europepmc.org/article/PMC/PMC13719417
AB - Reliable prediction of pharmaceutical solubility in supercritical carbon dioxide (scCO2) is complicated by strong compound-specific behavior and response values. This study investigated whether sparse measurements from a newly introduced drug can calibrate transferable machine-learning predictions for computational estimation of solubility in wide range of parameters. A curated workbook containing 1,618 observations for 60 drugs was modeled using molecular weight, melting point, temperature, and pressure. Solubility was transformed as log₁₀(S), and XGBoost, CatBoost, and multilayer perceptron regression were each tuned using tree-structured Parzen estimation, differential evolution, and particle swarm optimization. Evaluation used five-fold nested group cross-validation, with complete drugs held out from training and tuning. For each test drug, predictions were assessed before calibration and after 1, 3, or 5 randomly selected measurements, repeated ten times. Within the present evaluation framework, the lowest-log-MAE zero-shot configuration was XGBoost optimized by tree-structured Parzen estimation (log-MAE = 0.584; R² = −0.087; RMSE = 1.713 g/L), showing that the four descriptors were insufficient for reliable extrapolation to unseen compounds. Sparse calibration substantially improved transfer: the lowest-log-MAE one-shot result was obtained by CatBoost–particle swarm optimization (log-MAE = 0.418; R² = 0.166), whereas XGBoost–tree-structured Parzen estimation had the lowest log-MAE at three shots (log-MAE = 0.298; R² = 0.317) and at five shots (log-MAE = 0.251; R² = 0.452; RMSE = 1.223 g/L). Median absolute relative deviation decreased from 82.4% to 42.9% between zero and five shots. These results establish sparse, leakage-controlled calibration as a more defensible use case than unconditional prediction for unseen drugs, while identifying the need for richer molecular descriptors and external prospective validation. In practical terms, this few-shot strategy can reduce the experimental burden for a newly introduced drug by using a small screening set to calibrate a global model before mapping the remaining pressure-temperature domain.
SN - 2296-858X
ER -