A multi-objective RL pipeline with a pretrained LSTM generates covalent inhibitor candidates for EGFR and ACHE, rediscovering known inhibitors and spontaneously creating novel warhead motifs absent from training data but supported by independent literature.
On the Value of Using 3D Shape and Electrostatic Similarities in Deep Generative Methods.Journal of Chemical Information and Modeling, 62(6):1388–1398
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Multi-Objective Reinforcement Learning for Generating Covalent Inhibitor Candidates
A multi-objective RL pipeline with a pretrained LSTM generates covalent inhibitor candidates for EGFR and ACHE, rediscovering known inhibitors and spontaneously creating novel warhead motifs absent from training data but supported by independent literature.