Update and verify wang2012 wsas weights for leaprc.protein.ff14SB

[bieniekmateusz]

Temporarily chosen the consensus based on the atom type

[bieniekmateusz]

Hey @shunzhou , this is something for us to remember and take care of at some point. WSAS uses weights for each atom type: but there are new atom types.

Notes

2C as CB 
there is also CX, C0 and C8
CX is clearly CA, 
C8 is tricky: could be CG, CB, CD etc, but they're similar in json, only CG has a higher weight,
picking CB again
CO could be cd or cg, but i pick cd for now

[bieniekmateusz]

Hey @shunzhou and @apbhati1 ,

I have to finish a few things. If you know which paper it is that would be useful. I only know that from the file name this wsas-params-wang2012.json

Here is the configuration now with the new atom types introduced (copied based on others):
https://github.com/UCL-CCS/BAC2/blob/master/bac/analyse/wsas/data/wsas-params-wang2012.json

[bieniekmateusz]

More specifically, these are the updates:
6bc3da8#diff-7346c2aa7b9e580d6188f8608d6e7634
258c977#diff-7346c2aa7b9e580d6188f8608d6e7634

[shunzhou]

Hey @shunzhou , this is something for us to remember and take care of at some point. WSAS uses weights for each atom type: but there are new atom types.

Notes

2C as CB 
there is also CX, C0 and C8
CX is clearly CA, 
C8 is tricky: could be CG, CB, CD etc, but they're similar in json, only CG has a higher weight,
picking CB again
CO could be cd or cg, but i pick cd for now

We can check the sp hybridization of the carbon atoms:

2C, 3C, and C8 are all sp3 carbon. I think we can use the same weight as that in the 2012 paper.
CO is a sp2 carbon, so the weight for sp2 carbon in the paper should work.

[bieniekmateusz]

This is useful, thanks @shunzhou . However, it appears that they do change the weights regardless of hybridization. I'll search for the paper soon.