Complexes modeling
#2
by stolosa - opened
Hi!
I am trying to model the serine protease/inhibitor complex (1CGI) by giving as input both sequences separated by a "/" (sequence1/sequence2) and the model generates a structure for this complex but with an incorrect orientation of the monomers. Does anyone know how to use ESMFold with protein complexes?
Thanks :)
stolosa changed discussion title from Complexex modeling to Complexes modeling
You would insert a flexible (poly-G) linker and then remove it from the output. See the paper and see https://github.com/facebookresearch/esm/blob/main/esm/esmfold/v1/esmfold.py#L288