# Gen-COMPAS: Breaking Time Scales with Generative Sampling

Digest AI · Research · published 2026-09-09T00:00:00Z

Canonical: https://digestai.news/story/gen-compas-breaking-time-scales-with-generative-sampling

## Summary

Molecular transitions, including protein folding, are notoriously difficult to simulate due to their rarity and the computational demands of enhanced-sampling methods. A new framework called Gen-COMPAS has been introduced, which uses a denoising diffusion probabilistic model to reconstruct transition pathways without predefined collective variables. This approach allows for the identification of transition states and the generation of ensembles at nanosecond-to-submicrosecond scales, significantly reducing the computational cost compared to conventional methods. Applied to a range of systems, from a miniprotein to a pentameric ion channel, Gen-COMPAS has successfully recovered committors, transition states, and free-energy landscapes, providing a computationally tractable route to understanding biomolecular systems that have eluded conventional simulation techniques. This breakthrough could pave the way for more efficient and comprehensive studies of complex biological processes.

## Key points

- Gen-COMPAS uses a denoising diffusion model to reconstruct transition pathways.
- It identifies transition states without predefined collective variables.
- Applicable to a range of molecular systems, including a pentameric ion channel.

## Sources

1. [Breaking timescales with generative sampling of conformational transitions](https://nature.com/articles/s41586-026-11025-1) (Nature Machine Learning, 2026-09-09, primary source)

## Cite

Digest AI, "Gen-COMPAS: Breaking Time Scales with Generative Sampling", 9 September 2026, https://digestai.news/story/gen-compas-breaking-time-scales-with-generative-sampling

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/gen-compas-breaking-time-scales-with-generative-sampling.json