Guillermo Ignacio Benítez

This study shows that AlphaFold2 struggles to predict accurate 3D structures of proteins with conformational epistasis, as seen in the glucocorticoid receptor, but performs well for proteins without such epistasis like the mineralocorticoid receptor.

The paper identifies a conserved promiscuous reaction in serum albumins across mammals and suggests evolutionary adaptations in residues with unknown functions.

The paper evaluates 43 methods for predicting intrinsically disordered regions and binding regions in proteins, finding that deep learning methods outperform older physicochemical approaches but binding prediction remains challenging.