English

HEDGEHOG: Hierarchical Evaluation of Drug Generators Through Rigorous Filtration

Machine Learning 2026-07-14 v1 Software Engineering

Abstract

Generative molecular models can support early drug discovery by proposing new candidate compounds de novo. In practice, useful candidates must balance target-relevant activity, synthetic accessibility, physicochemical properties, and other multiparameter design constraints. However, metrics commonly used to evaluate molecular generators only weakly reflect whether the generated compounds are medicinally plausible and suitable for downstream computation. This can produce false positives in model evaluation, incorrect assumptions, and inefficient use of computational resources. We introduce HEDGEHOG, a unified six-stage filtration benchmark that is inspired by industrial hit identification workflows: (i) preprocessing; (ii) physicochemical descriptor screening; (iii) structural alerts and graph-sanity checks; (iv) synthesis feasibility; (v) docking and binding affinity estimation; and (vi) three-dimensional pose and interaction checks. We evaluate 23 molecular generators across three model classes under a standardized protocol. Across 230,000 generated molecules, only 0.65% of initial molecules survive all stages. Our results expose a central limitation of current molecular generators: molecules that appear acceptable under isolated criteria rarely satisfy medicinal chemistry, synthesis, docking, and 3D pose filters simultaneously.

Keywords

Cite

@article{arxiv.2607.13155,
  title  = {HEDGEHOG: Hierarchical Evaluation of Drug Generators Through Rigorous Filtration},
  author = {Daria A. Ryabchenko and Pavel Gurevich and Shamil Kadyrov and Daria Frolova and Kseniia Fedisheva and Sergei A. Nikolenko and Alexander Shapeev and Marina A. Pak},
  journal= {arXiv preprint arXiv:2607.13155},
  year   = {2026}
}

Comments

29 pages (including References and Appendix sections), 8 tables, 7 figures, 2 supplementary files