tech
A near-autonomous AI chemist improves a challenging reaction in medicinal chemistry
With Molecule.one’s Maria, GPT‑5.4 found a surprising additive boosting Chan-Lam Coupling yields for over 80% of tested substrates.
TL;DR
- GPT-5.4, connected to Molecule.one's Maria AI and high-throughput laboratory, improved yields for a challenging Chan-Lam coupling reaction.
- The AI identified TEMPO as a surprising additive that boosted yields for over 80% of tested sulfonamide and boronic acid substrates.
- The mean yield increased from 16.6% to 25.2%, with the proportion of reactions above 30% yield rising from 15.6% to 37.5%.
- Human chemists validated the microliter-scale results at bench scale, confirming yield increases for most substrate pairs.
- This advancement is crucial for drug discovery, as reliable synthesis is often a bottleneck, limiting the molecules scientists can test.
- The process involved AI proposing research ideas, designing experiments, and analyzing data, with human chemists providing oversight, corrections, and validation.