Hello Amber users,
I am trying to model a polymer containing different kinds of side-chains. Each side-chain is built individually, and are then assembled using the "seq" command in LEaP. I used the GAFF2 parameters to build my monomers, and RESP charges. Now, one of the side-chain that I want to simulate is ended by a glucose. So, I basically want to build a "modified glucose" which would be connected to a polymer backbone by a linker (one of the oxygen of the glucose is replaced by a linker).
My idea was to use GAFF2 parameters for the "linker" part and GLYCAM for the sugar. Concerning the parameters, I think that it should be OK if I add a frcmod file for the missing parameters (where "GAFF" atoms are linked to "GLYCAM" atoms), even though it would be necessary to check if these parameters are really good (mdgx could be used to reparameterize incorrect bonded terms). However, I am wondering if it is a good way to go concerning the partial charges. I see two possibilities here:
* Using RESP charges for the whole monomer, i.e., replacing the "GLYCAM" charges on the sugar by new, RESP charges. Problem: the GLYCAM parameters are probably dependent on the charges, so is it still correct to use the GLYCAM parameters with "non-GLYCAM" charges?
* Using RESP charges for the "GAFF" part, and keeping the "GLYCAM" charges for the sugar. Problem: the charges for GLYCAM have not been derived following the same methodology than RESP, some hydrogen atoms have a 0 charge, and the global charge of the monomer is not necessarily 0.
Another possibility that I see, which would be much simpler to implement, would be to use only GAFF2 to describe the monomer, as was done here for example (with GAFF): https://pubs.rsc.org/en/content/articlelanding/2018/cc/c8cc04818e#!divAbstract. Maybe GAFF2 can do an acceptable job to describe the glucose, which is a "simple" sugar? Again, some reparameterization could be done with mdgx, for example, to ensure viability of the parameters.
Thank you very much in advance,
David Dellemme.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Jan 29 2024 - 09:00:02 PST
