Integrating multiple protein language models using protein geometric graphs can dramatically improve the model performance for predicting the contacting residue pairs between interacting proteins.

Statistics on the frequencies of pi interactions in folded protein structures enable successful prediction of intrinsically disordered protein phase separation, with clear implications for a physical understanding of cellular organization.

Interactions in protein complexes can be predicted from evolutionary information from genomic sequences.

A new machine-learning toolbox unveils coevolutionary protein motifs related to structure, function, and phylogeny from sequence information only.

Diverse LXG toxins delivered by the Esx pathway function as mediators of contact-dependent interbacterial antagonism within and between Gram-positive bacterial species.

The number of contacts at the interface of a protein–protein complex, together with the properties of the surface, provides a simple, but well-performing predictor of binding affinity.

The active zone scaffold protein RIM-BP2 performs distinct functions in vesicle release at two hippocampal synapses, providing insights on how synapses express diversity in release properties.

Experimental determination of residue contacts from mutational data allows model discrimination and identification of in vivo functional conformations of proteins.

The physical interaction network encoded in the multi-domain protein native structure handles the trade-off between the fast, stable folding and the efficient, reliable function.

A combined theoretical and experimental study revealed the mechanistic and structural basis of mutational effects on allostery and provided insights into the intrinsic connection of intra- and inter-domain properties.