search

Junjie-Zhu / IDPFold

IDPFold test version
14 stars 2 forks source link

Question regarding the dataset used in the second stage of training. #2

Open hzk597955516 opened 3 weeks ago

hzk597955516 commented 3 weeks ago

Hello, nice work! I would like to ask how you obtained 77,600 conformations from 3,880 systems in IDRome?

Junjie-Zhu commented 2 weeks ago

Hi, Thank you for following our work!

We extracted 20 conformations from each trajectory first, and fixed the structures with PDBFixer. We then ran energy minimization with OpenMM to make sure the structures were physically reasonable. The script for this process refers to Amber relaxation in AlphaFold2, you can also find it here: https://github.com/Junjie-Zhu/Phanto-IDP/blob/main/Analysis/refine_openmm.py

hzk597955516 commented 2 weeks ago

extracted 20 conformations from each trajectory

Thank you very much for your reply! I would like to ask about the extraction strategy for the 20 conformations of each trajectory.

Junjie-Zhu commented 2 weeks ago

The 20 frames are extracted at equal interval from trajectories.