[gmx-users] per residue RMSF calculation - reg

Vigneshwar Ramakrishnan vigneshwar.ramakrishnan at gmail.com
Tue Mar 7 12:13:04 CET 2017

Dear all,

I am trying to understand the calculation behind RMSF per residue (that is
obtained by using the -res flag in the rmsf command).

I understand how per atom calculations are done. My initial thought was
that the RMSF per residue is the average over all the RMSF of the atoms in
the residue. That is, if a residue has n atoms (n1, n2, n3 ... n), then the
RMSF of that residue is computed as

RMSF_residue = sum(RMSF(n1) + RMSF(n2) + ... RMSF(n))/n

One can calculate RMSF(n1), RMSF (n2) can be calculated using the gmx_rmsf
command withOUT the -res flag.

The per residue RMSF is calculated by using the -res flag.

I tried to calculate the per residue RMSF using the above formula using the
data obtained for the atomic rmsf (calculated without the -res flag). The
values don't match.

Look forward to your help in understanding the calculation.

The details of the commands and the values are as follows:

g_rmsf_d -f HSA-MD-5-10ns-50ps.xtc -s HSAmd1.gro -o rmsf.xvg -res
g_rmsf_d -f HSA-MD-5-10ns-50ps.xtc -s HSAmd1.gro -o rmsf-atom

(selected protein group in both cases).

The atomic RMSF  values for one serine are as follows

    1   0.1881
    2   0.2140
    3   0.2138
    4   0.2216
    5   0.1661
    6   0.1665
    7   0.1897
    8   0.2021
    9   0.2077
   10   0.2104
   11   0.2261
   12   0.1485
   13   0.1877

The per residue RMSF for the above Serine residue is 0.1866

Thank you,

More information about the gromacs.org_gmx-users mailing list