[gmx-users] g_covar

[tangxuan]

Dear all,
When I am using g_covar to calculate the covariance matrix for alpha 
carbons in one subunit of my protein, I get a warning" WARNING number of 
atoms in tpx (929) and trajectory (293401) do not match". 929 is the 
alpha carbon number and 293401 is  whole system atom number. In the 
command, I used trr file and tpr file of the whole system. Do I have to 
get a new trr file and tpr file just for these alpha carbons? 
Some information  is shown below.
"Group     0 (      System) has 293401 elements
Group     1 (     Protein) has 75295 elements
Group     2 (   Protein-H) has 38096 elements
Group     3 (     C-alpha) has  4831 elements
Group     4 (    Backbone) has 14493 elements
Group     5 (   MainChain) has 19340 elements
Group     6 (MainChain+Cb) has 23750 elements
Group     7 ( MainChain+H) has 23963 elements
Group     8 (   SideChain) has 51332 elements
Group     9 ( SideChain-H) has 18756 elements
Group    10 ( Prot-Masses) has 75295 elements
Group    11 ( Non-Protein) has 218106 elements
Group    12 (         KCX) has   184 elements
.
.
Group   107 (chO_chV_&_C-alpha) has   929 elements
Select a group: 107
Selected 107: 'chO_chV_&_C-alpha'
Calculating the average structure ...
trn version: GMX_trn_file (single precision)
Reading frame       0 time 20000.000  
WARNING number of atoms in tpx (929) and trajectory (293401) do not match
Last frame       1000 time 30000.000  

Last frame       1000 time 30000.000   to ./#average.pdb.17#
Read 1001 frames

Trace of the covariance matrix: 2.34563 (nm^2)
100%
Diagonalizing ...

Sum of the eigenvalues: 2.34563 (nm^2)

Writing eigenvalues to eigenval.xvg

Writing reference, average structure & eigenvectors 1--1000 to eigenvec.trr
Wrote the log to covar.log"

Tang