# Compute RDF between COMs of two selections (or groups) using GROMACS

I have been using gromacs since years but I am always confused when I want to perform the following task: compute the RDF (radial distribution function) between the COMs (center of masses) of two selections or groups.

I am assuming I can define selections within the command line via -sel and -ref- options, however I am always confused if to include com there or in the -selrpos option.

Example: I have several residues and I want to compute the RDF between the COMs of the selection made by atoms C1 C2 C3 (in all residues). Possible syntax:

1$$gmx rdf -f ftraj.xtc -s ftraj.tpr -ref "com of name C1 C2 C3" -sel "com of name C1 C2 C3" -selrpos res_com 2$$ gmx rdf -f ftraj.xtc -s ftraj.tpr -ref "com of name C1 C2 C3" -sel "com of name C1 C2 C3"
3$$gmx rdf -f ftraj.xtc -s ftraj.tpr -ref "name C1 C2 C3" -sel "name C1 C2 C3" -selrpos res_com 4$$ gmx rdf -f ftraj.xtc -s ftraj.tpr -ref "name C1 C2 C3" -sel "name C1 C2 C3" -selrpos res_com -seltype res_com
5$gmx rdf -f ftraj.xtc -s ftraj.tpr -ref "name C1 C2 C3" -sel "name C1 C2 C3" -selrpos whole_res_com -seltype whole_res_com 1$ and 2$return zero everywhere, while 3$ peaks at atomic positions (not COMs). It seems only last commands (4$or 5$) are doing somehow what I am looking for but I am not sure what is the difference (and there is difference). So, further confusion is due to the option listed for selrpos: what would be the difference between res_com, whole_res_com, part_res_com and dyn_res_com? Also, the option seltype is only documented in last versions of gmx rdf.

Besides, I am assuming I could pre-define "C1 C2 C3" as a group via gmx make_ndx

Any clarification is welcome :)

• Did you figure this out? I can offer a bounty to help get it solved. Commented Jan 15, 2021 at 7:23
• I have found a solution, yes. I'll post it asap! Commented Jan 15, 2021 at 10:58