> I am currently facing a problem in deciding the partial charges on a
> certain subsections of a DNA. For example if I take a DNA seq of ATTC the
> first chunk would be assigned -0.3079 charge as per the amber library and
> the last chunk i.e. 3' end assigned -0.6921 charge. Ideally the 5' should
> have no charge as no phosphate group is there and 3' end should be -1
> charge as one phosphate group is present (as per the amber standard
> residues). Please suggest what is the reason behind this charge
> distribution and how to decide the charges on the terminal.
Yes, it is true that, if the 5' phosphate is absent as is often the case,
that we think that the "residue" should have no charge and that 3' has a
phosphate and should be -1. These are "formal" charges.
However, how do you define residue and influence of the phosphate on the
charge distribution? Or, why are you imposing formal charges on a
molecular system that is simply trying to approximate the overall charge
distribution as accurately as possible? Does the -1 sit entirely on the
phosphate?
For a monovalent ion, it is fairly easy to describe. For example Na+ has
a +1 charge, so obviously it should be +1. Maybe this approximation is
OK (although we know that likely this is not perfect since charge
redistributes, but I digress)...
OK, let's jump further and think about Mg2+; should it have a +2 charge
too? Well, depends on how we want to model. Given that Mg2+ is usually
hexa-hydrated, should we model it as Mg2+ (+2 charge) with neutral waters
or should we assume a bonded model and physically distribute the +2 charge
over the Mg and the 6 H20? We can do either. The former is easy and
allows water to exchange; the latter may more accurately reflect the
charge distribution, but does not allow waters to exchange. There are
many alternatives.
How about the phosphate, should we do:
OH OH
| |
O=P-O- or -O-P=O
| |
C C
or a mix? Logically, the mix is probably better. Should the -1 be only
on the P, only on the PO3- or distribute further?
OK, now for 5'-3'; where should we break the connection so that one side
is -1 and the other neutral? This will depend on the charge fit...
With the AMBER force field, others early decided that internal nucleotides
should be neutral. They then did best fits to reproduce the electrostatic
potential on the ends; it turned out, with a natural break at residue
boundaries the 3' end was slightly more negative than the 5' end and that
both, when added together added up to the required -1... That is the key.
The break point is arbitrary and why should we localize the -1 to a small
region?
(another way to think about it is by looking at the actual charges; you
will see that the phosphate is not -1 but that charge is distributed to
best fit the ESP from QM calculations)
Hope this helps. --tom
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Apr 23 2015 - 23:00:03 PDT
