# De novo drug design

3 related articles

Subscribe to updates

Journal of Chemical Information and Modeling Nov 9, 2026
Pharmacophore constraints go straight into 3D diffusion: PharmaDiff hits a 0.6990 exact-match rate without a target structure
PharmaDiff turns a 3D pharmacophore hypothesis into an atom-level subgraph and injects it as a condition into diffusion-based generation, producing 3D molecules that follow the prescribed spatial arrangement. Across 250 pharmacophore conditions drawn from the GEOM-Drugs test set it reaches a mean match score of 0.8964 and a perfect match rate of 0.6990, above PGMG, TransPharmer, REINVENT4, ShEPhERD and PhoreGen; its average docking scores on four targets also beat DiffSBDD, which needs the protein structure, though its PoseBusters pass rate is below ShEPhERD and PhoreGen.
PharmaDiff turns a 3D pharmacophore hypothesis into an atom-level subgraph and injects it as a condition into diffusion-based generation, producing 3D molecules that follow the prescribed spatial arrangement. Across 250 pharmacophore conditions drawn from the GEOM-Drugs test set it reaches a mean match score of 0.8964 and a perfect match rate of 0.6990, above PGMG, TransPharmer, REINVENT4, ShEPhERD and PhoreGen; its average docking scores on four targets also beat DiffSBDD, which needs the protein structure, though its PoseBusters pass rate is below ShEPhERD and PhoreGen.
AbbVie (United States)Carnegie Mellon University
Amira Alakhdar（First author）Newell Washburn（Corresponding author）
PharmaDiff Diffusion Model Pharmacophore modeling De novo drug design
bioRxiv Sep 18, 2026
TRACEDD: Tool-grounded Reasoning and Agentic Coordination for Explainable Drug Design
This work introduces TRACEDD, a tool-first multi-agent framework in which large language models serve as the reasoning and orchestration layer while domain-validated computational tools supply structural, chemical, pharmacological, and synthetic predictions, and it demonstrates an end-to-end JAK2 workflow spanning target validation, structure retrieval or AlphaFold prediction, druggable pocket identification, reinforcement-learning-based de novo molecular generation, lead optimization, ADMET and bioactivity evaluation, literature evidence integration, and retrosynthesis planning, with each decision linked to explicit tool invocations and intermediate evidence.
This work introduces TRACEDD, a tool-first multi-agent framework in which large language models serve as the reasoning and orchestration layer while domain-validated computational tools supply structural, chemical, pharmacological, and synthetic predictions, and it demonstrates an end-to-end JAK2 workflow spanning target validation, structure retrieval or AlphaFold prediction, druggable pocket identification, reinforcement-learning-based de novo molecular generation, lead optimization, ADMET and bioactivity evaluation, literature evidence integration, and retrosynthesis planning, with each decision linked to explicit tool invocations and intermediate evidence.
Sarveswara Rao Vangala（First author）Arijit Roy（Corresponding author）
TRACEDD Tool-first orchestration De novo drug design Retrosynthesis Planning De Novo Molecular Generation
Journal of Chemical Information and Modeling Sep 14, 2026
PockLigGPT: Pocket-Sequence-Conditioned Molecular Generation with GPTs and RL
This work introduces PockLigGPT, a GPT-based framework for ligand generation conditioned on the amino acid sequence of a protein pocket, trained in four stages (large-scale ZINC20 chemical pretraining, ChEMBL bioactivity-oriented adaptation, pocket-sequence-conditioned fine-tuning, and pocket-specific docking-guided reinforcement learning with AutoDock Vina-based rewards), achieving competitive docking-oriented performance under a standardized evaluation protocol while maintaining chemical plausibility and Lipinski-based drug-likeness, with docking studies on Alzheimer's disease-associated targets and token-level analyses supporting its utility for de novo drug design.
This work introduces PockLigGPT, a GPT-based framework for ligand generation conditioned on the amino acid sequence of a protein pocket, trained in four stages (large-scale ZINC20 chemical pretraining, ChEMBL bioactivity-oriented adaptation, pocket-sequence-conditioned fine-tuning, and pocket-specific docking-guided reinforcement learning with AutoDock Vina-based rewards), achieving competitive docking-oriented performance under a standardized evaluation protocol while maintaining chemical plausibility and Lipinski-based drug-likeness, with docking studies on Alzheimer's disease-associated targets and token-level analyses supporting its utility for de novo drug design.
Universidad Autónoma de Madrid Consejo Superior de Investigaciones Científicas
Pablo Varas Pardo（First author）
PockLigGPT De novo drug design Reinforcement Learning

You’re all caught up

Pharmacophore constraints go straight into 3D diffusion: PharmaDiff hits a 0.6990 exact-match rate without a target structure

TRACEDD: Tool-grounded Reasoning and Agentic Coordination for Explainable Drug Design

PockLigGPT: Pocket-Sequence-Conditioned Molecular Generation with GPTs and RL