@OpenAI: GPT-5.4 helped drive a medicinal chemistry project from literature review to a validated experimental result. Paired wi…
Summary
GPT-5.4, in collaboration with Molecule.one's Maria AI platform, autonomously drove a medicinal chemistry project from literature review to validated experimental result, proposing an unexpected improvement to a widely used reaction in drug discovery.
View Cached Full Text
Cached at: 06/17/26, 06:00 PM
GPT-5.4 helped drive a medicinal chemistry project from literature review to a validated experimental result.
Paired with https://t.co/gcDaph8b2B’s Maria AI and specialized lab, the model proposed an unexpected way to improve a widely used reaction in drug discovery. https://t.co/KmyBlHLX8y
Chemistry AI for Autonomous Discovery
Source: https://molecule.one/ Molecule.one’s Maria™ unites AI, Lab, and Data to power near-autonomous research for drug discovery & manufacturing. Commercially proven.
OpenAI and molecule.one collaboration:
Chan-Lam Synthesis Discovery OA1-M1-003
In a first for organic chemistry, Molecule.one & OpenAI have successfully put AI to work on an open-ended scientific problem. GPT-5.4 & Maria AI picked the research area, generated proposals, rated them, and ran the experiments in the Maria Lab. OpenAI post|Preprint
The Maria™ AI platform
### Maria AI Frontier chemistry models combined in an agentic framework to solve chemical problems and discover new chemistry
### Maria Lab The revolutionary, micro-liter scale HTE lab directed by Maria AI and tailored for her workflows
### Maria Data The industry leading, novel data set generated at Maria AI’s direction in her purpose-built lab
Press Highlights
Grace and Molecule.one to Transform Peptide Building Block Synthesis with AI and Automation December 8, 2025
Standard Industries Announces Molecule.one as Winner of $1 Million AI Challenge June 12, 2025
TechBio Spotlight – Molecule.one by Mason Victors October 15, 2024
CAS and Molecule.one Announce a Strategic Collaboration to Accelerate Drug Discovery August 11, 2023
Papers
- JCIM, 2021
Molecule Edit Graph Attention Network: Modeling Chemical Reactions as Sequences of Graph Edits 2021 - EFMC-ASMC 2025
10× Higher Hit Rates At 3–10× Lower Cost: Accelerating Ligand Discovery With ML-Guided Synthesis & Screening 2025 - NeurIPS Workshop AI4Science, 2025
Trustworthy Retrosynthesis: Mitigating Hallucinations with Reaction Plausibility Filtering and Retrieval-Augmented Scoring 2025 - NeurIPS Workshop AI4Science, 2025
Scaling High-Throughput Experimentation Unlocks Robust Reaction-Outcome Prediction 2025 - Journal of Cheminformatics, 2024
Relative molecule self-attention transformer 2024
Case studies
- Engineering Serendipity: Rapid Single-Iteration Hit Discovery using Plate Analogs in μSpaceM1 Case Study
First Application of AI-Driven Small Molecule Synthesis in a Non-Pharmaceutical Industry Case Study
Accelerating Hit-to-Lead Campaigns with molecule.one Case Study
Rapid, Single-Iteration Hit Identification for CLK1 using Direct-to-Biology Case Study
Posts
- How Did a Startup End Up Running World’s Largest Microliter-Scale HTE Reaction Campaign? October 31, 2025
- Trustworthy Retrosynthesis: How a Culture of Chemist-AI Collaboration Won $1M and Led to a Strategic Partnership with W.R. Grace December 11, 2025
Stay in touch, join our newsletter
Similar Articles
A near-autonomous AI chemist improves a challenging reaction in medicinal chemistry
OpenAI connected GPT-5.4 to an autonomous chemistry AI (Maria) to improve Chan-Lam coupling of primary sulfonamides, achieving significant yield improvements in medicinal chemistry reactions.
@OpenAI: Maria tested the idea across 10,080 reactions, and human chemists later validated representative results by hand. Under…
OpenAI and Molecule.one collaborated to have their AI systems (GPT-5.4 and Maria) autonomously select research areas, generate proposals, and run experiments in organic chemistry, achieving yield improvements for 88% of tested reactions — a first for AI-driven open-ended scientific discovery.
Medical research with GPT-5
OpenAI announces the application of GPT-5 to medical research, highlighting the model's capabilities in healthcare and scientific discovery domains.
Early experiments in accelerating science with GPT-5
OpenAI releases a research paper documenting early experiments where GPT-5 accelerates scientific discovery across mathematics, physics, biology, and other fields, including cases where the model helped identify disease mechanisms, solve open problems, and improve optimization algorithms in collaboration with leading universities and national laboratories.
Measuring AI’s capability to accelerate biological research
OpenAI demonstrates GPT-5's capability to accelerate biological research by autonomously optimizing a molecular cloning protocol in collaboration with Red Queen Bio, achieving a 79-fold improvement in cloning efficiency through novel enzymatic mechanisms. The work showcases AI's potential to support experimental iteration and empirical validation in wet lab settings while highlighting biosecurity considerations.