[gmx-users] g_order output is mirrored

[Justin A. Lemkul]

Igor Marques wrote:
> dear all,
> 
> i've been trying to use g_order to analyse the disorder of the chains of 
> POPC model membrane simulations, namely the SCD parameters.
> 
> one example of "desired" output is this:
> 5IGl.png 
> http://i.imm.io/5IGl.png - Performance of the general amber force field 
> in modeling aqueous POPC membrane bilayers by Balázs Jójárt Tamás A. 
> Martinek DOI: 10.1002/jcc.20748
> 
> 
> however, my output, when plotted, presents a looks like mirrored:
> deuter.png
> http://i.imm.io/5IGK.png
> 
> 
> my input index files start with the first carbon after the c=o in each 
> chain, as in the tutorial 
> - http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin/gmx-tutorials/membrane_protein/tutorial.html
> if started the other way around, the plot resembles published data...
> 
> a colleague tried the same process with a DMPC membrane model and his 
> plots are also mirrored.
> 
> our command line:
> g_order -f ../TRAJECTORY.xtc -s ../TOPOLOGY.tpr -n INDEX_CHAIN_sn1.ndx 
> -od OUTPUT_SCD.xvg
> 
> what are we not doing/doing wrong?
> 

A few questions:

1. What version of Gromacs is this?

2. Can you post your .ndx file?

3. What is your command for the sn-2 chain?  You cannot obtain unsaturated and 
saturated order parameters in the same command.  You must do them separately, 
and the index groups are different.  For instance, to get the order parameters 
around the unsaturation, your index file should contain only Cn-1, Cn, Cn+1, 
Cn+2, where Cn and Cn+1 are the C atoms joined by the double bond.  Then run 
g_order with the -unsat option.  These values can then replace the corresponding 
carbon positions in the g_order output of the whole sn-2 chain.  If done 
improperly, you can get weird results.

4. What are the simulation conditions?  Are you trying to replicate the 
published work?  Different simulations conditions will certainly impact the 
structural parameters, and insufficient sampling can give weird looking curves, 
as well.  All it might mean is that your data are not yet well-converged.

-Justin

> best regards and thanks in advance!
> 
>       Igor Marques
> 

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

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

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