Pyrite: Building and Benchmarking Protein-Ligand Pose Ranking Agents

Researchers submitted Pyrite, a pairwise protein-ligand pose-ranking benchmark of 1,500 pose pairs, to arXiv on 10 Nov 2026, reporting that two LLM-based agents built on a structured representation called PyriteRep beat expert chemists on the task. PyriteFeaturizer, which generates executable code to build features for logistic regression, reached 74% accuracy, and PyriteReasoner, which compares poses through text-based reasoning, reached 67%, while expert chemists using 3D visualization tools and physics-based scores performed near chance. Group Relative Policy Optimization (GRPO) improved open-weight models enough to match or surpass closed-source frontier models in the evaluations, and LLM-guided feature discovery generally outperformed direct reasoning.

Computer Science Machine Learning Submitted on 10 Nov 2026 Title:Pyrite: Building and Benchmarking Protein-Ligand Pose Ranking Agents View PDF http://arxiv.org/pdf/2611.14511v1 HTML experimental https://arxiv.org/html/2611.14511v1 Abstract:Predicting how a ligand binds to its target is central to small-molecule drug discovery. Modern cofolding models can generate diverse candidate poses, but reliably ranking these candidates remains challenging. We introduce Pyrite, a pairwise pose-ranking benchmark comprising 1,500 pose pairs across diverse protein-ligand complexes. Expert chemists equipped with 3D visualization tools and physics-based scores perform near chance, highlighting the difficulty of interpreting available chemical signals. To investigate automated ranking, we introduce PyriteRep, a structured representation of molecular geometry, protein-ligand interactions, and physicochemical scores, and develop two LLM-based agents: PyriteReasoner, which directly compares poses through text-based reasoning, and PyriteFeaturizer, which generates executable code to construct features for logistic regression, optionally incorporating features derived from LLM subcalls. These agents achieve up to 67% and 74% accuracy, respectively, exceeding observed expert performance. LLM-guided feature discovery generally outperforms direct reasoning, while Group Relative Policy Optimization GRPO improves open-weight models, enabling them to match or surpass closed-source frontier models in our evaluations. These results establish Pyrite as a challenging test of pose-ranking ability and highlight LLM-guided feature discovery as a promising approach to interpret and combine the outputs of computational chemistry tools. Submission history From: Megha Srivastava view email http://arxiv.org/show-email/0fe135fa/2611.14511 v1 Tue, 10 Nov 2026 18:56:32 UTC 5,909 KB Additional Features Current browse context: cs.LG References & Citations Loading... Bibliographic and Citation Tools Bibliographic Explorer What is the Explorer? https://info.arxiv.org/labs/showcase.html arxiv-bibliographic-explorer Connected Papers What is Connected Papers? https://www.connectedpapers.com/about Litmaps What is Litmaps? https://www.litmaps.co/ scite Smart Citations What are Smart Citations? https://www.scite.ai/ Code, Data and Media Associated with this Article alphaXiv What is alphaXiv? https://alphaxiv.org/ CatalyzeX Code Finder for Papers What is CatalyzeX? https://www.catalyzex.com DagsHub What is DagsHub? https://dagshub.com/ Gotit.pub What is GotitPub? http://gotit.pub/faq Hugging Face What is Huggingface? https://huggingface.co/huggingface ScienceCast What is ScienceCast? https://sciencecast.org/welcome Demos Recommenders and Search Tools Influence Flower What are Influence Flowers? https://influencemap.cmlab.dev/ CORE Recommender What is CORE? https://core.ac.uk/services/recommender IArxiv Recommender What is IArxiv? https://iarxiv.org/about arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs https://info.arxiv.org/labs/index.html .