2021-10-28 Meeting notes

General updates

• CD – Last meeting, we decided to push forward with ELF1 conformer generation.

• CD – Today I’m going to talk about differences I found in charge assignment with both toolkits when running with no optimization on MiniDrugBank.

Differences in charge assignment

• CD – We need to ensure that, when we plug the same conformer into the partial charge generation methods, we get the same partial charges out.

Open

• OM – What are the possible root causes?

  • Difference in AM1 implementation

  • Difference in BCCs are applied

  • Difference in chemical perception

  •  

  • CD – Can OE return just AM1-Mulliken charges?

    • LW – Yes: openeye: oequacpac.OEAM1Charges

  •  

• CD – Computing power issue – Analyzing 90 molecules took my computer 12 horus

  • OM – Do you have access to CU summit? Let’s get you on that.

  •  

  •  

• What do the bad molecules have in common?

  • About half have sulfonamides

  • About two thirds have sulfonyl (O=S=O)

  • All but one have indoles (many with additional Ns)

  • OM – We could use checkmol to list the functional groups that occur in the problematic parts of this dataset. SB made a python wrapper for it.

• How prevalent is this issue?

  • CD – about 9 out of 84 molecules I’ve tested so far. Minidrugbank has 350 molecules total, but I’ve only had time to test the first 84.

•  

Next steps

• Why does the data range end at exactly -1.0? Is something getting filtered out by the workflow? Maybe try putting cyanate in (C#N-)

• CD will make a minimal reproducing example

  • CD will send SMILES of good and bad mols to OM

• CD will try running just AM1 optimizations to see which deviations persist (if the parallel offset lines disappear, it’s a BCC issue)

• CD will turn off symmetrization to see if the vertical lines disappear.

  • For antechamber, could be antechamber’s -pl 0 or -eq 0 cli arguments (page 301 of http://ambermd.org/doc12/Amber21.pdf)

• JW will look into whether substituted indoles (indoles with extra Ns) are known cases of disagreement between aromaticity models/chemical representations.