search

luwei0917 / DynamicBind

repo for DynamicBind: Predicting ligand-specific protein-ligand complex structure with a deep equivariant generative model
MIT License
177 stars 38 forks source link

How to import ligand with "name" column in ligand.csv #21

Open stan1233 opened 4 months ago

stan1233 commented 4 months ago

Hi!

Thank you for your selfless open-source contribution, I have two questions.

How to give a name to a compound based on input ligand.csv instead of index+number in results. If I want to dock multiple proteins, is there a way to change the protein input to CSV?

Nobody-Zhang commented 4 months ago

Multiple Proteins...? Do you mean you have more than one chains? If you want to dock multiple proteins, you could just change the file which generates output files. @stan1233

stan1233 commented 4 months ago

Multiple Proteins...? Do you mean you have more than one chains? If you want to dock multiple proteins, you could just change the file which generates output files. @stan1233

What I want to express is how to assign a name to each ligand in the input ligand.csv? Also, I noticed that it seems possible in inference.py to submit multiple proteins and ligands for docking in one task. For instance, if I have 20 proteins and want to dock 200 ligands, how should I fill out the input.csv? Could you provide a template for the csv file?

Thank you for your reply, outstanding work!

aleff-ferreira commented 1 week ago

Hi,

To name your ligands in the ligand.csv file, you can include the SMILES string followed by the name or ID, separated by a space. Here’s an example of how I structure mine:

ligand
CCC1=C(C2=C(N1CC3=CC=CC=C3)C=CC=C2OCC(=O)O)C(=O)C(=O)N 155815
C1CC2=C(C1C(=O)N)C3=C(N2CC4=CC=CC=C4)C=CC=C3OCC(=O)O 9799156

This setup works for assigning specific names or IDs to each ligand.

If you’re looking to streamline the process of docking multiple ligands, I’ve put together a simple automation script. It can help with automating the ranking and selection of top ligands based on scoring metrics. Nothing fancy, but it might save you some time.

Feel free to check it out here: DBSA GitHub Repo.