I simulated a 600 a.a receptor containing two separate 300 a.a domains in complex with three different ligands. I want to calculate and compare the collective motion of each domain in these three different systems using gmx covar and gmx anaeig.
What I do is that defining my group of interest using gmx make_ndx, then fit least-square on one of the domains and calculate the covariance matrix for whole the receptor. Then, I use the achieved data to calculate PCA using gmx anaeig.
Is my approach physically relevant?
I would appreciate it if you would give me your opinion.
you can create a group in index file and select that particular index file for the calculations.
Najme Dehghanbanadaki you can use the following script to do a PCA analysis :
echo -e "3\n3" |gmx_mpi covar -f frame.xtc -s prod.tpr -ascii -n back.ndx
echo -e "3\n3" |gmx_mpi anaeig -s prod.tpr -f frame.xtc -v eigenvec.trr -eig eigenval.xvg -proj proj-ev1.xvg -extr ev1.pdb -rmsf rmsf-ev1.xvg -first 1 -last 1 -n back.ndx -2d
echo -e "3\n3" |gmx_mpi anaeig -s prod.tpr -f frame.xtc -v eigenvec.trr -eig eigenval.xvg -proj proj-ev2.xvg -extr ev2.pdb -rmsf rmsf-ev2.xvg -first 2 -last 2 -n back.ndx -2d
grep -v "^[#@]" proj-ev1.xvg | awk '{print $2}' > proj-ev11.dat
grep -v "^[#@]" proj-ev2.xvg | awk '{print $2}' > proj-ev12.dat
paste proj-ev11.dat proj-ev12.dat > nat_back.dat
- Badges
- Science topic