Robust Multi-Objective Reinforcement Learning Improves Joint Predicted-Potency and Drug-Likeness Yield in EGFR Molecule Generation
- 1Shanghai World Foreign Language Academy
- 2Department of Chemistry
- 3Duke University
Abstract
Optimizing a molecular generator against a single activity-model estimate can reward structures for which that estimate is unreliable. We extended our published ChemBERTa–reinforcement-learning (RL) workflow for human epidermal growth factor receptor (EGFR) with scaffold-separated activity modeling, empirically calibrated lower scores, bounded chemistry rewards, and matched-seed evaluation. From ChEMBL 37, we curated 6,476 unique structures with Reference/unspecified biochemical EGFR IC50 measurements; 4,491, 680, and 1,305 structures were assigned to scaffold-disjoint training, calibration, and test sets. The random-forest reward model achieved a test mean absolute error of 0.610 pIC50 units and R2 = 0.615. Uniform or potency-enriched masked-language-model continuation was crossed with three RL rewards, with three seeds and 600 generation attempts per combination. The locked LCB-versus-mean-only comparison changed GNN potency-screen yield by +0.28 percentage points (95% descriptive t interval,-1.10 to +1.66) across three seed-cluster means. In the pipeline-level comparison with the published-form baseline, Robust-LCB increased the GNN joint-screen yield in all six matched blocks; averaged across the two MLM arms, the yield rose from 0.72% to 1.22%. Two structures passed the full RF–GNN consensus funnel. These results support a pipeline-level increase in joint predicted-potency and drug-like yield, while the incremental contribution of the lower-score term alone remains uncertain. These comparisons describe computational variability; activity estimates for generated molecules require biochemical testing.
Related
- A Molecular Generation–Validation Workflow Combining Reversible Junction Tree Reinforcement Learning and Semi-empirical Quantum Chemistry: A Case Study on Molecular Solar Thermal Fuels
- Integrating DLVO Theory with Machine Learning for Predictive Toxicology and Multi-Objective Design of Quercetin-Loaded Nanocarriers
- Machine Learning of Band Gaps and Thermodynamic Stability in A2BB′X6 (X = F, Cl, Br, I) Double Perovskites with Uncertainty-Aware Multi-Objective Screening
- PCAN-PIE: A Machine Learning Tool for Prioritizing Prodrug Candidates by Predicted Bioactivation
- Transfer learning on universal interatomic potential embeddings improves generalization in structure-property defect models
BibTeX
@misc{guo2026robust,
title = {{Robust Multi-Objective Reinforcement Learning Improves Joint Predicted-Potency and Drug-Likeness Yield in EGFR Molecule Generation}},
author = {Yutong Guo and Xiao Huang},
howpublished = {ChemRxiv},
year = {2026},
month = sep,
doi = {10.26434/chemrxiv.15008490/v1},
url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15008490/v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Robust Multi-Objective Reinforcement Learning Improves Joint Predicted-Potency and Drug-Likeness Yield in EGFR Molecule Generation
AU - Yutong Guo
AU - Xiao Huang
PY - 2026
DA - 2026/09/08
PB - ChemRxiv
DO - 10.26434/chemrxiv.15008490/v1
UR - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15008490/v1
AB - Optimizing a molecular generator against a single activity-model estimate can reward structures for which that estimate is unreliable. We extended our published ChemBERTa–reinforcement-learning (RL) workflow for human epidermal growth factor receptor (EGFR) with scaffold-separated activity modeling, empirically calibrated lower scores, bounded chemistry rewards, and matched-seed evaluation. From ChEMBL 37, we curated 6,476 unique structures with Reference/unspecified biochemical EGFR IC50 measurements; 4,491, 680, and 1,305 structures were assigned to scaffold-disjoint training, calibration, and test sets. The random-forest reward model achieved a test mean absolute error of 0.610 pIC50 units and R2 = 0.615. Uniform or potency-enriched masked-language-model continuation was crossed with three RL rewards, with three seeds and 600 generation attempts per combination. The locked LCB-versus-mean-only comparison changed GNN potency-screen yield by +0.28 percentage points (95% descriptive t interval,-1.10 to +1.66) across three seed-cluster means. In the pipeline-level comparison with the published-form baseline, Robust-LCB increased the GNN joint-screen yield in all six matched blocks; averaged across the two MLM arms, the yield rose from 0.72% to 1.22%. Two structures passed the full RF–GNN consensus funnel. These results support a pipeline-level increase in joint predicted-potency and drug-like yield, while the incremental contribution of the lower-score term alone remains uncertain. These comparisons describe computational variability; activity estimates for generated molecules require biochemical testing.
ER -