Re: AMBER: Hydrogen bonding question

From: Thomas E. Cheatham, III <>
Date: Tue, 15 Mar 2005 08:59:00 -0700 (Mountain Standard Time)

> >I am trying to simulate a single strand of an oligonucleotide. This
> >results in a the strand folding onto itself in implicit solvent.
> >Then I tried to place it in explicit solvent and equilibrate the pressure
> >to 1 bar. Then I removed the restraints from the molecule and let it
> >move. Agin, this resulted in the strand folding onto itself.
> >
> do you have reason to think that this is not correct?
> Yes, I am trying to simulate a 12-mer PNA. Other simulations with PNA's,
> done by other people in the literature, seem to indicate that PNA's are
> not prone to hydrogen-bond onto itself. (Example of this can be seen
> here: JACS, 2001, 123, 7414-7422, Srikanta Sen, Lennart Nilsson) So based
> on this, I thought that PNA's are not likely to fold onto itself and
> hydrogen bond with nucleotides on the other end of the strand. If this is
> not the case, is this just an arteract of the program used in the paper
> mentioned above? And if this i the case, how can I be more certain of
> what to use and what is "more accurate"?

Are there experimental observations of the PNA single strand that suggest
it does not self-associate? Basing your intuition on a single set of
short calculations (with a different force field) is not fully warranted
unless there is distinct experimental data to back it up.

How to make it more accurate? Use RESP charges; characterize the dihedrals
to make sure the barriers are consistent with expectations (i.e. compare
what antechamber gives compared to the standard nucleic acid and protein
force field; perhaps characterize the PNA backbone residues with QM).
Compare your force field to that of others (Nilsson; Armitage; ...).

I will note that previously I ran a series of single stranded DNA
simulations (unpublished). If I started as a helix (i.e. take a double
helix and remove one strand) the structures are metastable on a 10-20 ns
time scale. If I randomly perturb the structure (by heating), the
structure will tend to fold back up towards an (imperfect) helix. This is
most notable with polyA (which as is known from experiment, adenine bases
alone will tend to stack). The metastability (even for polyT or polyG) is
of some concern since it suggests long time scales are needed to sample
the relevant substates.

I think the situtation you are trying to look at is going to be very
complicated. (1) We do not know if the force field is accurate (but let's
assume it is until proven otherwise), (2) there is a huge sampling problem
since likely the single strand can fold into a continuum of substates
(necessitating very long simulations or application of enhanced sampling
methods like replica-exchange), and (3) what experimental data do you have
to tell if what you observe in the simulations is "correct"? The
experimental data is fairly rich (albeit not at the atomistic structural
level) for single stranded DNA; I am not aware of such data for PNA or
other modified DNA's single stranded conformations...

Good luck.

\ Thomas E. Cheatham, III (Assistant Professor) College of Pharmacy, Depts of
| Medicinal Chemistry and of Pharmaceutics and Pharmaceutical Chemistry
| Adjunct Asst Prof of Bioengineering; Center for High Performance Computing
| University of Utah, 30 South 2000 East, Skaggs 201, Salt Lake City, UT 84112
| (801) 587-9652; FAX: (801) 585-9119
\ BPRP295A / INSCC 418
The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to
Received on Tue Mar 15 2005 - 16:53:00 PST
Custom Search