BioEmu is capable of generating thousands of statistically independent protein structures per hour on a single graphics card (GPU). Particularly noteworthy is the system's ability to detect complex and biologically relevant structural changes such as hidden binding pockets, movements of entire protein domains, or local unfolding. According to a recent press release from the FU Berlin, BioEmu can predict changes in protein stability with an accuracy comparable to that of laboratory analyses. 

According to the report, BioEmu offers new perspectives for drug design and could help increase the success rate of new drugs in clinical trials in the future.
The research was conducted entirely at Microsoft under the direction of Prof. Dr. Frank Noé from the Freie Universität Berlin. Significant contributions were also made by Prof. Dr. Cecilia Clementi, Einstein Professor of Physics at Freie Universität Berlin.

Original publication:
Lewis S, Hempel T, Jiménez-Luna J, Gastegger M, Xie Y, Foong AYK, Satorras VG, Abdin O, Veeling BS, Zaporozhets I, Chen Y, Yang S, Foster AE, Schneuing A, Nigam J, Barbero F, Stimper V, Campbell A, Yim J, Lienen M, Shi Y, Zheng S, Schulz H, Munir U, Sordillo R, Tomioka R, Clementi C, Noé F. (2025). Scalable emulation of protein equilibrium ensembles with generative deep learning. Science. 2025 Jul 10:eadv9817. doi: 10.1126/science.adv9817. Epub ahead of print. PMID: 40638710.

Source and further information:
https://www.fu-berlin.de/presse/informationen/fup/2025/fup_25_120-proteinmodellierung-bioemu-noe/index.html