[gmx-users] PCA uses average structure or reference structure

[sunita gupta]

hi all
I did MD studies of a dimer using NAMD and charmm forcefields.....now, in
order analyze dynamic motions PCA was done using the command g_covar of
GROMACS.....here i gave .trr and .pdb as input.....and selected backbone for
analysis.....the output obtained were average.pdb, average.trr....etc....
after that for analysing i used g_anaeig command.....providing average.trr
as input.....to get filter.pdb....(after loading it in VMD i could see the
global movement very clearly...)
my query is that " in a paper i read that PCA analysis does not make any
sense by using reference structure.....in fact average structure is the best
choice".I am confused whether what i did was correct or not......
and can anyone tell me how to use average structure in GROMACS


-- 
SUNITA GUPTA
Member Research Team
Lead Invent Technology
TBIU, IIT Delhi, India
Email- sunita at leadinvent.com
Ph- +9111 26581524 (Ex-6)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20080625/bc6c9fb1/attachment.html>