[gmx-users] Error

[Justin Lemkul]

On 10/10/15 9:52 AM, asasa qsqs wrote:
> Dear David van der Spoel,which plot must be check? I used this command "g_hbond -f md.xtc -s md.tpr -n index.ndx -num DrugwithWater.xvg" and i get a data point in the end of run of this command and i used it for report. I want to report this data point and it's error. what must i do?many thanks,

Posting the same or similar questions from multiple email addresses (and it 
appears, multiple names with similar questions) will not improve your odds of 
getting a faster answer. It actually looks more like spam.

You should read up on basic statistics.  The output of -num is a time series of 
the number of hydrogen bonds.  How might one estimate the error of this time 
series?  Are your saved frames of sufficient separation to be considered 
independent?  Have you done multiple runs over which you can average and obtain 
a meaningful standard deviation?  All of these are routine considerations in 
simulations.  GROMACS isn't going to magically output error estimates for any 
arbitrary quantity; sometimes you have to learn to do the statistical analysis 
yourself.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================