AMBER Archive (2005)

Subject: Re: AMBER: RESP charges of RNA nucleotides with 3' and 5' ends

From: FyD (fyd_at_u-picardie.fr)
Date: Thu Jul 07 2005 - 10:54:00 CDT


Quoting Ilyas Yildirim <yildirim_at_pas.rochester.edu>:

> I have been trying to understand how Cieplak et al. 95 paper calculated
> the charges for RNA residues, and have some questions on this. I have
> read Qing Zhang's "Procedure of Computing Partial Charges of
> Carcinogen-Modified B-Deoxynucleotides" paper, in order to understand
> the RESP procedure. I will appreciate if you could help me to get some
> answer on my questions. My main concern is how to get the RESP charges
> for 3' and 5' end of a strand. I would like to write what I have
> understood from that paper, and have your ideas whether I understood it
> right or not.
>
> Let's say that we want to get the RESP charges for cytosine residue,
> and call it "RC" (As it is called in xleap this way). We want to calculate
> the charges for RC, RC3 and RC5. RC3 and RC5 are the same residues when
> they are at 3' or 5' end of a strand. (As it is used in xleap this way)
>
> So, in order to find the RESP charges for these 3 structures we would
> follow the following procedure (Thats what I have extracted from Qing
> Zhang and Cieplak et al. 95 papers):
>
> 1. Start with the nucleoside structure (here it is cytidine). Use an ab
> initio program to calculate the ESP grid points.

(i) geometry optimization & (ii) MEP computation

> 2. Start another structure; dimethylphosphate (DMP). Use an ab initio
> program and get the ESP grid points.

Yes from conformation gauche, gauche

> 3. Use the program 'resp' to calculate the charges.
>
> This is what I have extracted from these both papers. But the 3rd step
> looks pretty confusing. In Qing Zhang's paper, at step 5, there are 2
> stages defined.
>
> On the first stage, it says that we have to define the charge
> constraints. As far as I understood from that, it means that if we want to
> calculate the RESP charges for RC5, this constraint is I + II = 0 ( I, II,
> III, and IV are defined below). If we want to calculate the charges for
> RC3, this constraint is III + IV = 0. If we want to calculate the RESP
> charges for RC, the constraints are I + II = III + IV = 0. After defining
> the charge constraint, on the first fit, we equivalence the O1P and O2P
> atoms, and the H atoms in NH2 group. And a hyperbolic restraint with a
> force constant a=0.0005 is used. I hope I am right till now.
>
> Here, a couple of questions arises:
>
> 1. What is the meaning of 'equivalencing the atoms'? I know that at the
> end, they will have the same charges, but this term 'equivalencing' is
> used in Cieplak et al. 95. page 1366, such that all sugar atoms are
> equivalenced. How can someone equivalence the sugar atoms of 4 different
> residues? Does that mean that they do a multiple molecule fit to these 4
> different residues and equivalence the sugar atoms?

See the R.E.D.-II manual on the web site. Charge equivalencing (eq) can be
perfomed on many things...
- Atoms in the same molecule (for instance H in CH3 or NH2)
- Atoms in different mol. orientations
- Atoms in diff mol. conformations
- Atoms in diff molecules
In diff nucleosides, charge eq. is performed on all the sugar atoms accept C1' &
H1' (the ones which connect to the base)

> 2. And how can we include the charge constraints in the resp input file? I
> have read the AMBER manual's RESP part, but without an example and the
> meanings of each lines in the resp input file, it is very hard to
> understand what each line means. (Especially I am confused after the 7th
> area of resp input file definitions)

Apply multi-orientation and or multi-conformation RESP fit with R.E.D.-I or R.E.
D.-II and you will get it...

After the 1st RESP section (Z, eq) jump two lines and start a new section to add
inter-molecular charge eq... See RESP manual.

If you want to add also inter-molecular charge restraints jump a unique line...

> On the second stage, the CH2 and CH3 are refit while the rest of the
> charges are kept. And a hyperbolic restraint with a force constant of
> a=0.001 is used. And the H atoms of CH2 and CH3 are equivalenced.

As it is done for CH2CH2OH in the R.E.D.-I tutorial

> Then, I have tried to analyze the sample case in Qing Zhang's paper. In
> the first fit, in the 2nd section of the resp input file, it is defined
> that we are doing a multiple molecule fit (here it is 2). On the 6th area,
> the atomic numbers and ivaries are defined. There are 3 different
> definitions for the ivary: 0, n and -99. When it is -99, this means that
> "Dont change the charge anymore". But when it is defined as '0' or 'n',
> the charges are varied. Here another term is used; 'center'. What does
> this mean? Is this just used when we equivalence some of the atoms?

-99 is the same meaning that -1. Please see the R.E.D.-II manual all is
explained. We used -1 in R.E.D. You can use -99 if you prefer ;-)

> After the 6th area, I could not understand the rest of the lines. Here is
> the copy of those lines used in this sample:
>
> 6 0.000
> 1 6 1 14 2 2 2 3 2 4 2 5
> 6 0.000
> 1 1 1 9 2 10 2 11 2 12 2 13

This inter-molecular charge restraints
6 atoms are involved with a total charge = 0
The atoms are the following: molecule 1 atom 6 molecule 1 atom 14 molecule 2
atom 2. etc... See the RESP manual...

> The 7th area is for charge constraints and the 7.1th area is for the list
> of the atoms that will be constrained according to this charge
> constraints. Now, here, I would assume that "6 0.0000" is the charge
> constraint and the next line is the atoms that will be constrained on the
> first molecule, while the 3rd line is again the charge constraint and the
> 4th line is the atoms in the second molecule (DMP) that will be
> constrained.

See above

> As I said in the beginning of my email, my main concern is to find the
> RESP charges of a residue when it is on 3' or 5' end. If there is any
> automated procedure to find the RESP charges for a 3' and 5' end, that
> will be very helpful. I have been using RED to calculate the RESP charges
> for an arbitrary residue, but when it is 3' or 5' end, I am not sure how
> to use RED to calculate the charges (because the total charge of the
> molecule is not anymore 0 or -1). I will greatly appreciate any
> ideas/references. Thanks in advance.

- You run R.E.D.-II for each molecule.
- Combine the espot files in a unque espot file
- Combine the RESP input1/2 of each molecule adding inter-mooecular charge
equivalencing and inter-molecular charge restraints
are run the follwing two aliases:
[fduprado_at_griffy ~]$ which resp1
resp1: aliased to resp -O -i input1 -e espot -o output1 -p punch1 -q qout -t
qout1 -w qwts -s esout
[fduprado_at_griffy ~]$ which resp2
resp2: aliased to resp -O -i input2 -e espot -o output2 -p punch2 -q qout1 -t
qout2 -w qwts -s esout
[fduprado_at_griffy ~]$

This is automatically perfomed in R.E.D.-III

> The structure of DMP is as follows:

Regards, Francois

-- 
  F.-Y. Dupradeau
DMAG EA 3901 & Faculte de Pharmacie, Amiens, France
The Scripps Research Institute, La Jolla, CA, USA
       --
http://www.u-picardie.fr/labo/lbpd/FyD.htm
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu