true/false properties in .mae file not working

[arosale4]

Whether I use 0 or 1 for b_cs_first_match_only, or b_cs_both_enantiomers I can not seem to reduce the number of structures produced. I have even gone into merge.py and set first_match_only to True, and it would still match every instance of a pattern. I'm trying every version of schrodinger that we have available to see if that is a problem. I have also noticed that the pattern PD-C2=C2 will also match PD.C2=C2.

[ericchansen]

Any updates on this?

[ericchansen]

@arosale4 Do your recent changes solve this issue?

[arosale4]

Thanks for drawing my attention to this. The recent changes may help with it, but indirectly. I do think we fixed this in 68da398?

[arosale4]

Or at least we addressed the issue of reducing matches from first_match_only() to one.

[ericchansen]

Closing this issue as part of the Q2MM modernization to an open-source SDK.

Why: This issue was specific to Schrodinger's .mae file format and merge.py structure manipulation -- both of which depended on the proprietary schrodinger.structure API. The modern Q2MM SDK no longer uses .mae files or Schrodinger's structure merging; molecules are built from standard open-source formats (Mol2, Gaussian .fchk, QCElemental, raw coordinate arrays).

What the modern SDK does instead: The Q2MMMolecule class uses RDKit for SMILES-based construction and handles bond/angle detection internally. Torsion and improper handling uses topology-based atom type matching rather than Schrodinger atom indices, avoiding the index-remapping problems described here entirely.

Thanks @arosale4 for the original report -- the underlying insight about atom identity vs. index was sound, and the new architecture addresses it structurally.