19 August 2026

Anthropic's Claude runs protein design experiments, beating typical success rates

• Anthropic tested Claude models (Mythos Preview and Opus 4.8) on designing minibinders, small proteins that block target proteins, a foundation for drug development.
• Of 1,320 designs Claude created against 15 protein targets, 354 actually bound in lab tests, a 26.8 percent success rate versus the typical 10 to 15 percent in the field.
• Claude did not build new protein prediction software. It installed and orchestrated existing open-source tools through a 16,000-word instruction set, automating what usually requires human expert decisions and days of work.
• A second experiment showed Claude could interpret raw scientific instrument data (nuclear magnetic resonance and mass spectrometry files) and produce analysis in under 25 minutes, sometimes discovering file formats on its own.
• Anthropic acknowledges no human expert comparison exists and designs were only tested for binding, not biological function or actual structure. Independent peer review of results is still pending.

Reported by The Decoder