AMBER Archive (2005)

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

From: Qing Zhang (qingzhang_nyu_at_yahoo.com)
Date: Thu Jul 07 2005 - 03:30:14 CDT


Dear Ilyas,

Please see my answers in your message body below.

> 1. Now, after running
> 'resp' we will have RESP charges for the nucleotide
> and DMP molecules. How
> are we going to decide what the charges for DX is?
> Are we going to use the
> O3', O5', P, O1P, O2P charges from DMP molecule as
> the charges for missing
> atoms in the DX molecule?

Yes, because the O3¡¯ group and the O5¡¯ groups in the
nucleotide are cancelled out by the two CH3 groups of
DMP in the two constraints. In the RESP result, you
will see two sets of partial charges, one from the
nucleotide and the other from DMP. You will have to
manually pick up the right charges for these five
atoms.

> 2. The above procedure is done for finding DX
> charges. If we want to find
> the DX3' or DX5', how should we define the resp
> input file? I would
> expect to use just one constraint (as it is stated
> in Qing Zhang's
> paper).

Yes, just use one constraint for each case. For
example (DX5), you constrain group I and II (they are
cancelled out) in RESP, but keep the O5¡¯ group of the
nucleotide after RESP.

> 3. In the above protocol, on the second resp stage,
> CH2 and CH3 parts of
> the nucleoside is refit, not the CH3 part of the DMP
> molecule. Why? Is it
> because we have already constrained those CH3 parts
> in the first resp fit?
> If that is so, how should we treat the DX3' and DX5'
> cases? In both these
> cases, we will use just one constraint. Does that
> mean that we should
> refit the CH3 part of the DMP molecule, too, which
> was not used on the
> first resp fit?

The CH3 groups in DMP will always be discarded after
RESP, there is no need to refit them.

> 4. And again, after the 'resp' calculation, how
> should we define the
> missing charges in either DX3' and DX5'?

First of all, the constrained part (I or III) in the
nucleotide and the two CH3 groups in DMP should be
discarded. Second, the unconstrained part in the
nucleotide is kept. Third, the phosphate group (P,
O1P, O2P) is used in DX3 but not in DX5, because the
latter does not have this group.

> 5. I still do not understand the 8th and 9th parts
> of the 'resp' input
> file; intermolecular charge constraints and multiple
> molecule atom
> equivalencing.

Intermolecular charge constraints is like the
constraints between the nucleotide and DMP, such as
I+II=0 or III+IV=0. It is similar to the 7th part.

Regarding multiple molecule atom equivalencing (also
the first question in your previous email), it is used
to equivalence the charges of the DNA/RNA backbone
atoms (please check Cieplak et al. 1995 for sure)
between A, T(U), G, C. That is how AMBER partial
charges were developed for the DNA/RNA (see Cieplak et
al. 1995). Thus, our protocol (without this
equivalencing) will give these atoms the partial
charges slightly different from AMBER. To study a
specific molecule (instead of developing the AMBER
force field), this equivalencing is usually not used.

Setting up RESP runs is time consuming. I also hope
this process can be automatic in the future.

Sincerely,

Qing

======================================
Qing Zhang, Ph.D.
Research Associate
Molecular Graphics Laboratory
Department of Molecular Biology, MB-5
The Scripps Research Institute
10550 North Torrey Pines Road
La Jolla, CA 92037-1000
Tel: (858) 784-2202
Fax: (858) 784-2860
Email: qzhang_at_scripps.edu
======================================

                
____________________________________________________
Sell on Yahoo! Auctions – no fees. Bid on great items.
http://auctions.yahoo.com/
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu