jwohlwend
Is smiles format the only solution to add the ligand?
I don't know why, but after using boltz-1 for prediction, a cyclohexane structure in my ligand turned into a benzene ring. Since I'm not very familiar with the SMILES format, I tried using SMILES from different websites, but the results were similar. Then I tried using a "SMILES to Structure" software to see if it could convert to the correct structure, and it showed that the SMILES format was fine. What could be the issue here? Are there any other ways to upload the ligand?
You can input CCD and sdf as well. I have the same situation and requested the review to the developer.
You can input CCD and sdf as well. I have the same situation and requested the review to the developer.
How to input sdf file?
You can input CCD and sdf as well. I have the same situation and requested the review to the developer.
How to input sdf file?
put your smiles into chemdraw and then save it as a sdf file
probably connected to https://github.com/jwohlwend/boltz/issues/195#issuecomment-2699631160 and https://github.com/jwohlwend/boltz/issues/195#issuecomment-2699663049. Still not 100% certain that my answer is correct, just crosslinking for more visibility on both issues. Does using SDF or CCD instead of SMILES fix the issue?
You can input CCD and sdf as well. I have the same situation and requested the review to the developer.
How to input sdf file?
put your smiles into chemdraw and then save it as a sdf file
Thank you for your reply. I have the ligand sdf file downloaded from Pubchem. Should I just copy that into yaml file and do the simulation?
You can input CCD and sdf as well. I have the same situation and requested the review to the developer.
How to input sdf file?
`structures:
- protein: id: A pdb_path: /app/data/pdb/XXXX.pdb
- ligand: id: LIG smiles: Cc1nc`
I write the Yaml file like this. Each pdb has one ligand.
probably connected to #195 (comment) and #195 (comment). Still not 100% certain that my answer is correct, just crosslinking for more visibility on both issues. Does using SDF or CCD instead of SMILES fix the issue?
Unfortunately, I can't fix it. I think that the Boltz-1 determine the energy minimization for the ligand. Despite of using SDF, CCD, SMILES, the Boltz-1 convert to internal ligand information, therefore, the ligand type isn't important.
probably connected to #195 (comment) and #195 (comment). Still not 100% certain that my answer is correct, just crosslinking for more visibility on both issues. Does using SDF or CCD instead of SMILES fix the issue?
Unfortunately, I can't fix it. I think that the Boltz-1 determine the energy minimization for the ligand. Despite of using SDF, CCD, SMILES, the Boltz-1 convert to internal ligand information, therefore, the ligand type isn't important.
So no matter the format of ligand, it will be modified during simulation, right?