[gmx-users] counting the number of water molecules surrounding proteins

[sang eun jee]

Dear Gromacs Users

Hello.

I have a question about how to count the number of water molecules
surrounding protiens.
I have tried two methods and the results from two methods were different.

The first one is use g_trjorder -f structure.xtc -s structure.tpr -n
index.ndx -nshell water_in_0.22.xvg -r 0.22
I have chose first reference group as protein and second group water.
Using this method, I could calculate the number of water molecules in 0.22
nm as time from protein.

The second method is using mindist.
At first I have calcualted minimum distance from water to protien within
0.5 nm using g_mindist
g_mindist -s structure.tpr -f structure.xtc -n index.ndx -d 0.50
-respertime -od od.xvg -or mindrest.xvg
Here I chose the first reference group water and second reference group
protein. As long as I understand, mindrest.xvg include minimum distance
data of the water-protein per each water molecule.

And then using grep, I have extracted the number of water molecules in the
range of r<0.22

grep -v '[#|@|S]' mindistres.xvg | awk '{a1=0;;for(i=2;i<=NF;i++)
if($i<'0.22') a1++;print $1,a1}' > water_count.xvg
Then I could obtain the number of water molecules in 0.22 nm as a function
of time.

When I got time-averaged values of the number of water molecules within
0.22 nm, the value from first method is different from second method. I got
162 water molecules from first method, while I got 222 water molecules from
second one.
Does anyone have experience in this method?

Thanks,

Sang Eun Jee

Post Doctoral Researcher
School of Materials Science and Engineering
Georgia Institute of Technology