Satya Nadella announces BioEmu: Microsoft’s AI to fast-track drug discovery

Microsoft has introduced BioEmu, an innovative artificial intelligence platform designed to accelerate the understanding of protein behavior within the human body. Traditionally, this process involves complex computer simulations that can take years to yield results. During the announcement on social media, Microsoft Chairman and CEO Satya Nadella emphasized the importance of comprehending protein motion for advancing biology and drug discovery, stating, "Today we’re introducing BioEmu, an AI system that emulates the structural ensembles proteins adopt, delivering insights in hours that would otherwise require years of simulation." Developed by Microsoft Research’s AI for Science team, BioEmu is capable of predicting the various shapes and movements proteins undergo as they carry out their functions in living organisms. This capability is crucial for understanding disease mechanisms and designing new drugs that target specific proteins. In a comprehensive breakdown shared online, Microsoft Research revealed that BioEmu version 1.1 closely aligns with real-world experimental data on protein stability, achieving prediction errors of less than 1 kcal/mol and maintaining strong correlation scores above 0.6 across extensive test datasets. This AI was trained on over 200 milliseconds of molecular dynamics simulations, data from more than 500,000 protein stability experiments, and a wealth of structural information. Unlike conventional methods that demand extensive GPU resources and time, BioEmu can perform the same simulations in just hours, drastically reducing both computational time and costs. One of its remarkable features is the ability to identify subtle changes in protein structure, including the elusive “cryptic” binding pockets—hidden areas on proteins that could be potential targets for new therapies. "BioEmu can emulate equilibrium distributions of millisecond-timescale molecular dynamics simulations at vastly accelerated speeds," noted Microsoft Research. The AI also predicts significant movements, such as large domain shifts and local unfolding, which are often critical to protein functionality. The findings have been published in the journal Science, positioning BioEmu as a generative deep learning model capable of replicating the structural ensembles of proteins both in laboratory conditions and the human body. These protein ensembles play a vital role in understanding protein functions, particularly as many proteins continuously transition between different forms. The launch of BioEmu is anticipated to revolutionize areas like drug development, disease research, and synthetic biology, potentially enabling scientists to discover and test new therapies at an unprecedented pace.