On 7/21/23 4:08 AM, majort--- via modeller_usage wrote: > Hi I'm modeling protein with ligands (using the "." options in the > sequence file). However, in the final models the contacts between > protein side chains and ligands are too close (the template and > target have differently sized ligands; the template target ligand has > a smaller ligand). Is there a way to avoid this, like giving less > weight to the homology terms and greater weight to the standard FF > terms?
When using the '.' syntax Modeller copies the ligand and its protein-ligand interactions directly from template to target so - as you've seen - it won't work very well if the environment differs from template to target. Your main options here would be
1. Find a different template that has the ligand in the same environment as the target.
2. Define force field terms for the ligand and try to model it flexibly rather than as a rigid body.
3. Take your Modeller-generated model and try to refine the protein-ligand interactions with a good MD package.
Ben Webb, Modeller Caretaker