Principal component analysis (PCA) for studying mutation effect

GROMACS version: 2023.1
GROMACS modification: No

Hi everyone,
I need to investigate the effects of a single-point mutation on my protein structure. I would like to run a free energy landscape analysis but first I need to compute PCA and I was starting with this command:

gmx covar -f md_noPBC.xtc -s md.tpr -o eigenval.xvg -v eigenvec.trr -tu ns

but I don’t know which group should be selected because the only examples I’ve found are about protein-ligand systems where are selected C-alpha and ligand groups. Is the selection of C-alpha and protein groups correct for studying the mutation effect on the protein structure? I’m asking because I’m not sure if it’s correct, and trying it’s particularly expensive considering that my simulation is 1 μs long and my protein has 1046 amino acids. After that, I was thinking about this command to analyze eigenvectors

gmx anaeig -v eigenvec.trr ­-f md_noPBC.xtc ­-eig eigenval.xvg ­-s md.tpr ­-first ­-last ­-2d 2dproj.xvg -comp eigcomp.xvg -rmsf eigrmsf.xvg -tu ns

In this case, is it more correct (or also convenient considering how long is my simulation) if I specify the first and last eigenvector to analyze? How should I understand their values? Is it a piece of information I can retrieve from the eigenvec.trr derived from gmx covar?

Thank you very much in advance,
Martina