SAMPLing OA system parametrization

[Steboss]

Hi everybody,

We were analysing the source of discrepancy we had between SAMPLing and SAMPL binding free energy predictions for the OA system (both OA-G3 and OA-G6 ). We found out that there is a mismatch between the OA host's charges used for the SAMPL and the ones adopted in SAMPLing. To parametrize the OA and TEMOA hosts for SAMPL, we have used SAMPL5 input files (OAH and OAMe), retrieving frcmod and off files, necessary to create AMBER type topology and coordinate files.

I would like to know how the OA SAMPLing host was parametrized, in order to shed light on SOMD predictions issue.

I attached a csv file, where you can find the charges for OA host system's atoms computed from the input files present here, and SAMPL5 OA input file. (the column order is: atom name, GROMACS charges, GROMACS_old charges, CHARMM charges, AMBER charges, SAMPL5 charges) charges.zip

Thank you,

Stefano

[davidlmobley]

Tagging @andrrizzi

(@Steboss - I think SAMPLing was a fairly independent setup of exactly what Andrea planned to run, not necessarily based on what was deposited here. The thought was that everyone would be running with exactly the SAMPLing input files, so it was not necessarily important to ensure match with the SAMPL files. But Andrea should be able to shed more light on exactly what he did.)

[andrrizzi]

Hi, @Steboss. The charges of all the molecules were determined using the OpenEye AM1-BCC scheme (see here for more details on the setup protocol). If the SAMPL5 charges were found with antechamber, we expect them to be slightly different.

[Steboss]

Hi @davidlmobley @andrrizzi Thanks for your replies. I do not know how the SAMPL5 were parametrized. @davidlmobley could you retrieve this information? According to my file, it seems there is a huge charge difference between some atoms in the OA molecule. This may lead to some interesting considerations, based on the charge methods employed.

[davidlmobley]

I don't know either, @Steboss -- it was Gilson lab folks for SAMPL5 I think. Probably Jane Yin...

[davidlmobley]

(Do you want to provide example files?)

[jmichel80]

@steboss can we repeat the calculations using the same input coordinates and protocol as used in SAMPLing but with the topologies used in SAMPL6 ? That should help clarify what is going on.

Julien

Sent from my iPhone

On 14 Mar 2018, at 20:41, David L. Mobley notifications@github.com<mailto:notifications@github.com> wrote:

(Do you want to provide example files?)

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/MobleyLab/SAMPL6/issues/50#issuecomment-373167237, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ALLd5EOhu_GywRxMsW0B-0Gx5oQCQNgZks5teYBigaJpZM4SqJ1G.

The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.

[Steboss]

@jmichel80 yes, I think this is a good starting point. @davidlmobley I can provide example files, but I think it is better to test a bit more the simulations at first. Thank you

Stefano

[davidlmobley]

@Steboss is this resolved, can we close the issue? Or is there more you need?

[Steboss]

Hi David, I think it’s definitely solved, so we can close it. Thanks very much