search

deepmodeling / Uni-Mol

Official Repository for the Uni-Mol Series Methods
MIT License
726 stars 127 forks source link

Ligand Preparation / Result Issue with Docking V2 #225

Open denizkavi opened 6 months ago

denizkavi commented 6 months ago

When running Uni-Mol Docking V2 via the demo.sh script on a custom protein-ligand, every part of the process seems to complete successfully, but the result file for the ligand is different than the input ligand. Do I need to prep the ligand in a certain way to resolve this?

image

joonseonyoon commented 6 months ago

I also have a similar issue, but not like that much of above image. Bond lengths are shortened with unfavorable angles. Can I find the docking v1 checkpoint file? Thanks for helps.

ZhouGengmo commented 6 months ago

You can try turning on the --steric-clash-fix parameter in demo.py. And is it convenient for you to share this sdf file? we would like to test it locally.

ZhouGengmo commented 6 months ago

I also have a similar issue, but not like that much of above image. Bond lengths are shortened with unfavorable angles. Can I find the docking v1 checkpoint file? Thanks for helps.

docking v1 checkpoint is here. The name is ‘Protein-ligand binding pose prediction’.

EasperOne commented 1 week ago

Although also have this issue when I run docking myself for some reasons, the result I got from the bohrium lab service is great. Any idea how to fix this?