Research map / Probabilistic and scientific modeling / Molecules and proteins
Refine Drugs, Don’t Complete Them: Uniform-Source Discrete Flows for Fragment-Based Drug Discovery
Benno Kaech, Luis Wyss, Karsten Borgwardt, Gianvito Grasso
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use uniform-source discrete flow with finer time granularity h (ablation) | InVirtuoGen · de novo generation | quality and diversity | Improved |
| replace standard discrete flow sampling with the proposed refinement sampling (ablation) | InVirtuoGen · de novo generation | quality-diversity pareto | Improved |
| increase time granularity under standard DFM sampling (ablation) | InVirtuoGen · de novo generation | quality-diversity pareto | No clear effect |
| remove 1/(1-t^2) time-weighting of the loss (ablation) | InVirtuoGen · de novo generation | quality-diversity pareto | Improved |
| add PPO to the genetic algorithm (ablation) | InVirtuoGen · PMO prescreen | sum AUC-top10: 17.758 -> 18.893 | Improved |
| add mutation of best molecules (ablation) | InVirtuoGen · PMO prescreen | sum AUC-top10: 17.831 -> 18.893 | Improved |
| add experience replay without prescreening (ablation) | InVirtuoGen · PMO | sum AUC-top10: 16.131 -> 16.676 | Improved |
| add peak-seeker bandit for sequence length (ablation) | InVirtuoGen · PMO prescreen | sum AUC-top10: 18.782 -> 18.893 | Improved |
| start sampling at t_start=0.2 instead of 0 (ablation) | InVirtuoGen · PMO prescreen | sum AUC-top10: 18.836 vs 18.893 | Mixed |
| prescreen ZINC250k before optimization | InVirtuoGen · PMO | sum AUC-top10: 16.676 -> 18.993 | Improved |
| enforce fragments by overwriting positions each step | InVirtuoGen · fragment-constrained generation | validity: 71.22 vs SAFE-GPT 85.00 | Hurt |
| InVirtuoGen versus GenMol on fragment tasks | InVirtuoGen · fragment-constrained generation | quality: 21.06 -> 29.55 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- GenMol — GenMol: A Drug Discovery Generalist with Discrete Diffusion (ICML 2025) · better on 5 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- GenMol: A Drug Discovery Generalist with Discrete Diffusion (ICML 2025 · similarity 0.88)
- Generative Flows on Synthetic Pathway for Drug Design (ICLR 2025 · similarity 0.87)
- Drug Discovery with Dynamic Goal-aware Fragments (ICML 2024 · similarity 0.87)
- Active Flow Expansion for Out-of-Distribution Discovery: from Theory to Molecules (NeurIPS 2026 · similarity 0.87)
- TFG-Flow: Training-free Guidance in Multimodal Generative Flow (ICLR 2025 · similarity 0.86)
- Controllable 3D Molecular Generation for Structure-Based Drug Design Through Bayesian Flow Networks and Gradient Integration (NeurIPS 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
