Hi Jesper,
As far as I know, this is what happens with the quasi-harmonic entropy
approximation. I believe it's documented in the literature, and there's
actually an expression for deriving the asymptotic limit of the entropy
using this approach.
HTH,
Jason
2011/5/16 Jesper Sørensen <lists.jsx.dk>
> Hi,
>
> So I am trying to calculate the entropy of a system that I have simulated
> for 100ns, and also to check it for convergence. It seems that the entropy
> increases (becomes more negative) the more simulation-data is included in
> the calculation, but it increases in a systematic way that does not make
> sense to me...
>
> I have calculated the entropy using MMPBSA.py with PTRAJ option.
> I have calculated the entropy from the simulation in different chunks to
> check the convergence....
> # 10 ns chunks
> 0-10ns
> 10-20ns
> 20-30ns
> ...
> 90-100ns
>
> # 20 ns chunks
> 0-20ns
> 10-30ns
> ...
> 80-100ns
>
> #....
>
> # 100ns chunks
> 0-100ns
>
>
> For the 10ns chunks the results end at around -225 kcal/mole
> For the 20ns chunks the results end at around -390 kcal/mole
> For the 40ns chunks the results end at around -700 kcal/mole
> For the 100ns chunk the results end at around -1400 kcal/mole
>
> Can this be right? - should the number systematically increase with the
> simulation length? It is only the vibrational part that varies
> significantly.
>
> The numbers surprise me a bit. As an example here are the results for the
> 0-100 ns chunk (using 500 frames):
> Translational Rotational Vibrational Total
> Complex: 17.0998 16.3344 1973.1877 2006.6222
> Receptor: 17.0524 16.2936 1965.0449 1998.3912
> Ligand: 14.4737 14.3574 1428.4709 1457.3017
>
> DELTA S: -14.4263 -14.3166 -1420.3282 -1449.0707
>
> The difference in vibrational entropy for the complex and receptor compared
> to the ligand seems WAY too small, especially compared to values that I get
> from a quick calculation using MMPBSA.PY with NAB for the entropy (although
> here I have only used 50 frames - in the interest of time).
>
> Complex:
> Entropy Term Average Std. Dev.
> -----------------------------------------------------------
> Translational: 17.0998 0.0000
> Rotational: 17.6933 0.0073
> Vibrational: 3716.0164 6.8456
> Total: 3750.8096 6.8495
>
> Receptor:
> Entropy Term Average Std. Dev.
> -----------------------------------------------------------
> Translational: 17.0524 0.0000
> Rotational: 17.6272 0.0129
> Vibrational: 3528.2633 7.5486
> Total: 3562.9430 7.5557
>
> Ligand:
> Entropy Term Average Std. Dev.
> -----------------------------------------------------------
> Translational: 14.4737 0.0000
> Rotational: 13.9148 0.1747
> Vibrational: 213.2078 5.1222
> Total: 241.5962 5.2772
>
> DELTA S total= -53.7296 +/- 10.0718
>
>
> I hope you can help me understand where the systematic error (if it is an
> error) comes in...
>
> Best,
> Jesper
>
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon May 16 2011 - 14:00:03 PDT
