AMBER Archive (2008)

Subject: Re: AMBER: charge derivation for modified amino acids

From: FyD (fyd_at_q4md-forcefieldtools.org)
Date: Wed Feb 27 2008 - 00:35:10 CST

Quoting Denzil Bernard <db4amber_at_gmail.com>:

> I am trying to generate charges for some modified amino acids using REDII
> and I had some questions regarding the application of constraints in the
> resp fitting procedure.

I would strongly suggest you to use R.E.D.-III instead of R.E.D.-II as
intra-molecular charge constraints (INTRA-MCC keyword has to be set in
the .p2n file), inter-molecular charge constraints (INTER-MCC keyword
has to be set in the .p2n file) and inter-molecular charge
equivalencing (INTER-MEQ keyword has to be set in the .p2n file) are
now automatically handled in R.E.D.-III.

> Is it sufficient to apply the constraint to the terminal ACE and NME group
> charges to zero with constraints on the amino acid residue NH and CO atoms
> set to standard amber values?

- You first set INTRA-MCC to zero for the ACE and NME blocking groups.

For this, did you look at the R.E.D. tutorial ?
http://q4md-forcefieldtools.org/Tutorial/Tutorial-1.php#10
See Scheme 1: for a central fragment fragment, two INTRA-MCC are set to zero.

Example: TABLE 12:
REMARK INTRA-MCC 0.0 | 1 2 3 4 5 6 | Remove
REMARK INTRA-MCC 0.0 | 20 21 22 23 24 25 | Rem

http://q4md-forcefieldtools.org/Tutorial/Tutorial-1.php#13
Scheme 2: for a terminal aminoacid fragment, one INTRA-MCC and one
INTER-MCC are set to zero.

- Then, if you look at the the NH and CO charges of the hydrophobic
amino-acids in the original AMBER force field topology database, you
will see that these charges are indentical. In this case, you can set
up additional INTRA-MCC keywords (each of them set to a particular
value) to follow what has been done in the past.

Please see http://q4md-forcefieldtools.org/Tutorial/Tutorial-1.php#10

Example TABLE 12:
REMARK INTRA-MCC -.4157 | 7 | Keep

> When is it necessary to set the the CO and NH atoms of the ACE and NME
> groups to standard amber values, and how does one do that while retaining
> group neutrality.

The NH and CO of ACE and NME do not need to be constraint to a
particular value. You might decide to constraint NH and CO of the
amino acid you are interested in, but not those of the ACE and NME
blocking groups.

> Is the second stage of charge fitting applied to both methyl and methylene
> groups or just for methyls.

In the second stage the charges of the methyl groups belonging to the
NME and ACE group are _not_ recomputed as these groupements are
removed (case of a central fragment). This allows to slightly improve
the RRMS value of the fit. On the contrary, the charges of all other
methyl or methylene groups (belonging to the amino acid) are obviously
recomputed in the second stage.

Please see once again the tutorial:
http://q4md-forcefieldtools.org/Tutorial/Tutorial-1.php#10
TABLE 13 - Job no 16 & edit the .p2n file:
http://q4md-forcefieldtools.org/Tutorial/5-Charges-2conf-RBRA-4-INTRA-MCC/Mol_red1.p2n
Read the REMARK about the charge equivalencing procedure applied to this case:

REMARK The methyl hydrogens of ACE and NME are not equivalenced to limit
REMARK the number of charge constraints. Since the ACE and NME residues
REMARK are removed from dimethylalanine dipeptide in the charge derivation
REMARK of the AIB fragment, the charge values of the ACE and NME methyl
REMARK groups do not need to be recomputed in the second RESP stage.

See TABLE 14 to download the data (the whole R.E.D.-III run)
corresponding to the JOb no 16.
http://q4md-forcefieldtools.org/Tutorial/5-Charges-2conf-RBRA-4-INTRA-MCC.tar.bz2

regards, Francois

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu