PharmAgents2: Towards Autonomous End-to-End Drug Discovery
Abstract
Drug discovery requires continuous scientific decision-making across a vast research space. We introduce PharmAgents2, an autonomous system for end-to-end drug discovery that combines knowledge-enhanced decision-making with organized large-scale exploration. Paper2Skill extracts transferable scientific decision knowledge from public sources and, together with expert-developed skills, supports research decisions. Self-Scaling Discovery expands exploration through parallel agents inheriting research context and decision-making capabilities across targets, scientific hypotheses, and discovery strategies. We introduce PharmAgents-Bench v0.1, comprising 121 questions across nine decision scenarios and three end-to-end tasks assessed with expert-defined process rubrics. PharmAgents2 outperforms evaluated baselines in most decision categories and achieves the highest process scores on all three end-to-end tasks. In three real-world studies, autonomous investigations generated findings subsequently supported by human-executed experiments. For LpxH, PharmAgents2 hypothesized a previously unreported allosteric site and screened compounds against it, identifying a novel inhibitor scaffold with experimentally confirmed inhibitory activity and binding. For JAK1, it prioritized a covalent inhibition strategy and nominated compounds with experimentally demonstrated activity. Starting from a known bioactive molecule, Self-Scaling Discovery explored candidate targets across the proteome and identified a new binding target, with the interaction subsequently confirmed experimentally. PharmAgents2 represents a prototype of a virtual pharmaceutical company, marking a step toward autonomous end-to-end drug discovery.
Figures & tables
| Task | Decision scenario | Evaluation metric |
| Binding-site assessment | Locate a binding pocket to guide structure-based ligand design. | Residue accuracy |
| Target hypothesis prioritization | Prioritize target hypotheses for a compound with an unexplained phenotype. | Known-target hit@50 |
| Molecular edit prioritization | Choose where and how to modify a parent structure during optimization. | Top-1 accuracy |
| Development candidate selection | Compare candidate molecules as a whole and select one to advance into development. | Selection accuracy |
| Molecular optimization outcome prediction | Anticipate property trade-offs during optimization toward a development candidate. | Property-direction score |
| Experimental affinity ranking | Prioritize candidates for testing by expected affinity. | Pairwise ordering accuracy |
| Case | Discovery narrative | Evidence status |
| LpxH | Allosteric Site Exploration and Functional Assessment Agent-Designed Discovery Plan Candidate Molecules | 63 candidates purchased and tested; AIPher-L1 identified with supporting evidence from enzyme inhibition and biophysical binding assays. Binding site identification is ongoing, with preliminary evidence supporting the proposed mechanism. |
| JAK1 | Known Covalent Site Binding-Pocket Exploration Candidate Molecules | 34 candidates purchased and tested; AIPher-J1 exhibited inhibitory activity comparable to that of the positive control. Three additional compounds achieved greater than 50% inhibition in both functional pathways. |
| Secondary-target discovery | Secondary-Target Hypothesis Parallel Target Exploration Ranked Target List and Research Reports | Exploration across 2,607 proteins produced 452 candidate dossiers and three target nominations, with Target X ranked first. Subsequent SPR experiments support binding of AIPher-X to Target X ( ). |
Drug Discovery
Drug discovery identifies, designs, and tests molecules that can safely affect disease biology.
See all 178 papers