[gmx-users] Strange behavior from g_sas

[Terry]

Hi, all,

I have four trajectories which I want to analyze Surface on. So I run 

for i in 1 2 3 4; do echo -e " non-Water \n Protein " | g_sas -f $i.xtc -s $i.tpr -o $i.area.xvg -n index.ndx -q $i.pdb ; done

But the strange thing is, g_sas analyzed the first three trajectories, but gave a "segmentation fault for the last one. 

The only difference for these trjs is the number of water molecules, hence the box size(I used NPT.) . Then I doubt maybe the atom name I used can't be put in pdb file (some of them are 5 letters long, and I used gro file since the very beginning ). I deleted the -q flag and run g_sas again, seg fault again.

After searching, I did this:
source /usr/local/gromacs/bin/GMXRC
Then g_sas worked fine( without the -q flag).

But I've already sourced GMXRC.bash.

Any clue?

Terry