search

nrbennet / dl_binder_design

MIT License
240 stars 57 forks source link

More than two chains? #30

Open okanders opened 1 year ago

okanders commented 1 year ago

Hi, I was wondering if this implementation has the architecture to handle more than two chains: in the case that I would like to filter my binders against a oligomer target? I feel like it would be difficult to properly model the inter chain relationships. Thanks so much and would love to talk more about this!

okanders commented 1 year ago

In the RFDiffusion filtering problem, how were you successful in recapitulating many chains with this program: specifically for the functional motif scaffolds or the double motifs that require multi-chain folding? Even for theoretical scaffolds of enzyme activation sites?

KoningCheems commented 8 months ago

I ran into the same problem :( Is this something that could be worked around with af initial guess? Given how it can predict the complex (binder+target) where target is composed of multiple chains merged into one chain with breaks. With initial guess AF then theoretically the output is not going to be the spaghetti mess?