2024-05-28 AMI/LW Check-in

General updates and discussions

• Ab Initio fit

  • AMI – weighted energies based on difference from QM, but something is overwritten. Waiting on benchmarks to see

  • LW – out of curiosity how long did it take?

  • AMI – ~2 days on 64 cores, normally takes about a week for a valence fit

• Compared YAMMBS vs old benchmark suite

  • Previously saw that overall distribution matched well

  • Compared molecule for molecule, and wound up getting pretty different molecules. e.g. 500 conformers with substantial differences, e.g. 50 kcal/mol ddE between conformers that have similar geometric metrics

  • Sulfurs are overrepresented but hard to say

  • Looking at ICRMSDs – most things tell the same story as before, except impropers

  • Comparing YAMMBS to old benchmarks

• Looking at 4-coordinate sulfur data

  • Target number of molecules?

    • Could start with 100 very different molecules, try besmarts and see how the splitting happens

• Roadmap: https://docs.google.com/presentation/d/1CYhur-N-M31WKa41VyGZa-3vlV6sinWepod2V-tumNI/edit#slide=id.g2c8f740d84f_2_22

• MS – on lipid parameters, our phosphorous/phosphate parameters might not be up to scratch