The London-based AI research firm DeepMind has introduced AlphaFold-Multimer, a model that can predict the structure of multi-chain protein complexes with increased accuracy.
The recent AlphaFold model can accurately predict many single protein chains. However, in many cases, the prediction of multi-chain protein complexes remains a challenge. Compared to the input-adapted single-chain AlphaFold, AlphaFold-Multimer significantly increased the accuracy of predicted multimeric interfaces while delivering high intra-chain accuracy. As the formation of protein complexes often lays the foundation of biological processes, DeepMind researchers hope this work will accelerate the research on complex folds, such as RNA and DNA molecules.
Protein complex prediction with AlphaFold-Multimer is on bioRxiv.