[gmx-users] about g_wham command of gromacs

[David van der Spoel]

On 01/09/15 23:09, 凌未风 wrote:
> Hello:
>
> I have a strange problem when I use g_wham to generate free energy profiles based on umbrella sampling (US) results.
>
> I have about ~30 windows of US simulations, 3 microseconds for each window, saved every 10 ps in the pullf.xvg files. When I try to generate the energy profiles, g_wham could not read the data in pull.xvg files after ~ 2.5 microseconds. When I use "g_wham -b 2500000", the program gives an error of no data, when I use "g_wham -b 0", it could generate reasonable profiles, but when I check the result, I found g_wham only reads data before ~2.5 microseconds.
> I double checked my pullf.xvg files, I am pretty sure there are "enough"  data in the files, and there is no break in the data. I also asked other colleagues in my lab as well as collaborators to help to check my data. We could not find where is wrong.
>
> I also tried the following things:
>
> 1, devided the time by 1000 in the pullf.xvg file, so the number of the time is not so large. It did not work.
> 2, calculate distances between groups based on trajectories, then change the format to "pdo". In one case, it worked, but in the other cases, it did not work.
>
> I am wondering if any one has the same problem. Did I do something wrong or this is a bug of g_wham?
>
> example of data is available upon request. Any help is highly appreciate.
>
> Thanks very much.
> Ruo-Xu Gu
> University of Calgary
>
This may have to do with an integer overflow. A fix is underway (but you 
can easily test it yourself, see details here 
https://gerrit.gromacs.org/#/c/5016/2).


-- 
David van der Spoel, Ph.D., Professor of Biology
Dept. of Cell & Molec. Biol., Uppsala University.
Box 596, 75124 Uppsala, Sweden. Phone:	+46184714205.
spoel at xray.bmc.uu.se    http://folding.bmc.uu.se