search

HannesStark / EquiBind

EquiBind: geometric deep learning for fast predictions of the 3D structure in which a small molecule binds to a protein
MIT License
469 stars 110 forks source link

Conformations are being sampled even with use_rdkit_coords=False #61

Closed jadolfbr closed 1 year ago

jadolfbr commented 1 year ago

The output ligand is significantly changing conformation even with use_rdkit_coords=False

This seems strange and is behavior that we do not want for some use cases. Is there anyway to force the input structure as the structure to bind?

HannesStark commented 1 year ago

Yes, this is what the rigid_redocking model does. There the ligand is treated as a rigid body and its conformation is not changed at all. In the config file you can specify the path rigid_redocking instead of flexible_self_docking.

Please let me know if you need more information to resolve your issue or if you can already get it running with that!

jadolfbr commented 1 year ago

Thanks for the info! This is very helpful!

On Sun, Sep 18, 2022 at 10:43 AM Hannes Stärk @.***> wrote:

Closed #61 https://github.com/HannesStark/EquiBind/issues/61 as completed.

— Reply to this email directly, view it on GitHub https://github.com/HannesStark/EquiBind/issues/61#event-7409029268, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAZDHRHFCFRXUD7JV55W2C3V64TBBANCNFSM6AAAAAAQNZI3VQ . You are receiving this because you authored the thread.Message ID: @.***>