Hi Vlad,
> In our PLoS Comput. Biol. study (Cojocaru et al 2011) with the GAFF
> force field for POPC published by Jojart and Martinek we used surface
> tension values ranging between 40 and 60 dyn/cm and found 50 and 60 to
> provide reasonable lipid properties for the time of the simulations
> (tens of ns) ... However, this was in NAMD and I don't know if the
> surface tension is the same implemented in AMBER 12 and NAMD ..
The implementation of the surface tension 'should' be the same. However,
that does not guarantee the actual implementation of GAFF in NAMD is the
same - getting AMBER and NAMD to match takes a lot of work. (see for example
http://ambermd.org/namd/namd_amber.html)
But... That's not the point here. The point is, to make sure we are all on
the SAME page, the original question concerned what surface tension term to
use for the "LIPID11" force field and NOT the GAFF force field. These are
DIFFERENT force fields and yet everybody here seems to be referring to what
value they used for GAFF. Ugh!!! There is no reason to assume the value
needed will be the same.
> I am curious about the references to the experimental values in Ross's
> plots .. We tried to reproduce an experimental value of approx. 68
> based
> on a recent study (don't remember now the reference and I am traveling
> right now)
If you look at the original plot I sent you will see there are two
experimental lines. The solid blue one is the one at 68 that you are
referring to. This is
(77) Kučerka, N.; Tristram-Nagle, S.; Nagle, J. F. J. Membr. Biol. 2005,
208, 193-202.
Which was based just on Xray. There is a second newer value based on a
combination of X-ray and neutron scattering:
(70) Kučerka, N.; Nieh, M.-P.; Katsaras, J. Biochim. Biopys. Acta 2011,
1808, 2761-2771.
Which is the dashed black line in that plot at around 64.5. The main point
being that one needs to consider that experiment can vary wildly and often
has much larger true errors on the values that the actual error bars
represent.
The value of 17 dyn/cm for the surface tension for POPC with "LIPID11" gives
you an area per lipid close to the newer experimental value. A value of 20
dyn/cm gives you a surface tension close to the original experimental value.
Note we have also compared to a number of other experimental properties such
as deuterium order parameters which imply a value of 17 dyn/cm is
appropriate. This info will all be in the manuscript once we manage to get
it published.
Something else to note, it can take MANY nanoseconds, on the order of 50 to
100 at least for a decent sized lipid bilayer (128 lipids) for the area per
lipid to converge. Many of the earlier papers that looked at using the GAFF
force field did not run their simulations long enough to actually present
converged results. This is possibly part of the reason why such a high
surface tensions are suggested. They also used small bilayers of 96 or less
lipids in some cases and thus some of the results seen might be affected by
system size artifacts.
All the best
Ross
/\
\/
|\oss Walker
---------------------------------------------------------
| Assistant Research Professor |
| San Diego Supercomputer Center |
| Adjunct Assistant Professor |
| Dept. of Chemistry and Biochemistry |
| University of California San Diego |
| NVIDIA Fellow |
|
http://www.rosswalker.co.uk |
http://www.wmd-lab.org/ |
| Tel: +1 858 822 0854 | EMail:- ross.rosswalker.co.uk |
---------------------------------------------------------
Note: Electronic Mail is not secure, has no guarantee of delivery, may not
be read every day, and should not be used for urgent or sensitive issues.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sat Jun 16 2012 - 09:30:02 PDT
