Does the coarse-grained representation remove double/triple bonds in ligands?

[ballaneypranav]

Hi, congratulations on the excellent work! The publication was highly engaging, and the results are very promising. I have a question about how DynamicBind processes ligands, specifically regarding its coarse-grained representation. Does this representation eliminate double or triple bonds?

I tried to run it on the test files (1qg8). While the input SMILES has double bonds, the output has only single bonds. This was confusing because the figures 3-5 in the publications show double bonds in the DynamicBind-predicted ligands. Could you please clarify what might be causing this issue?

Here's the command I ran:
python run_single_protein_inference.py data/origin-1qg8.pdb data/1qg8_input.csv --header test ...

2D representation of input SMILES and output ligand opened on Schrodinger Maestro:

Output complex:

Thanks for your help!

[luwei0917]

double and triple bonds should be conserved. I feel it might be a outputting issue related with pdb format. maybe checkout the predicted sdf file.