AmoAi

Leveraging physics-informed deep learning algorithms, we seek to understand evolutionary signatures from protein sequences, to use them to annotate protei functions, and to address some of the most complex and interesting challenges in biology.

---

How to identify evolutionary signatures?

Understanding the co-evolution of amino acid pairs within a protein is invaluable for applications such as structure prediction, mutation design, and protein engineering. Here, we present a method that harnesses estimated co-evolution information to pinpoint networks of residues, termed 'residue communities', which play a pivotal role in protein function.