search

TinkerTools / tinker9

Tinker9: Next Generation of Tinker with GPU Support
Other
48 stars 26 forks source link

Charge transfer energy does not agree between two versions of code #214

Open leucinw opened 1 year ago

leucinw commented 1 year ago

Hi @zhi-wang , I noted that my dynamic9 jobs failed with ele-lambda being 0 with the code of Feb 9. Error message is Induced dipole not converged. I compared a previous version (Jan 13), this one ran dynamic and finished. I ran analyze9 on two versions and I can see the Charge Transfer energy is different by 10 kcal/mol so I am thinking that the new code did not scale down CT energy. Could you help us solve the issue? For your test, here are the files I used: https://biomol.bme.utexas.edu/~liuchw/Tinker9_DMS/

zhi-wang commented 1 year ago

@roseane-reis Would you add some test cases for these situations? Thanks!