update logP analysis with 2 reference calculations

samplchallenges / SAMPL6

Challenge inputs, details, and results for the SAMPL6 series of challenges

https://samplchallenges.github.io

MIT License

52 stars 32 forks source link

update logP analysis with 2 reference calculations #79

Closed MehtapIsik closed 5 years ago

MehtapIsik commented 5 years ago

This PR adds 2 reference calculations to logP challenge analysis:

zxcvb - GAFF + tip3p
zaqws - GAFF + tip3pFB

It also corrects a file naming mistake of one of the blind submissions (vzgyt-1559-logP-shin-1.csv) which was interrupting the analysis script.

MehtapIsik commented 5 years ago

I like the idea of using separate submission IDs for references. We can use capital letters even to distinguish better: REF01, REF02, .... I had used previously nbXXX scheme for non-blind pKa change predictions. Which one is preferable?

We can make plots of reference and blind submissions separately too, but I think we should keep the plots that show everything together as well.

MehtapIsik commented 5 years ago

I prefer that we update reference submissions in a new PR instead of reverting this. Is that acceptable?

davidlmobley commented 5 years ago

Yes, a new PR is fine. And I'm enthusiastic about uppercase letters, REF01 etc sound good.

MehtapIsik commented 5 years ago

OK. @bergazin and I will update submission IDs accordingly.