Different conformers generated for the same amide molecule when different SMILES are provided

Dear developer,

I’m using EMC to build a liquid box for N-Methylacetamide, CC(=O)NC. I noticed that almost all molecules are in the cis conformation in the built box. After some debugging, I found that EMC generates different conformers when different SMILES strings are provided, even though they represent the exact same molecule:

mol0   CC(=O)N([H])C -> All trans conformer

mol0   CC(=O)N(C)[H] -> All cis conformer

Since EMC samples conformations using the Monte Carlo method, I would expect the final conformations to obey a Boltzmann distribution and therefore be irrelevant to the specific format of the SMILES string. Did I miss some settings to properly sample this?

I have attached the esh, top, prm files and the generated pdb, out files for your reference.

Thank you for your time and help!

amide-trans-cis.zip (12.5 KB)