AMBER Archive (2009)

Subject: RE: [AMBER] re: essential dynamics

From: Niel Henriksen (niel.henriksen_at_utah.edu)
Date: Wed May 13 2009 - 13:28:22 CDT

Hi Chunliyan,

I'm not an expert with IED. But I've done some tinkering . . .
I took a look at your input files. I think the problem is that you ran PCA analysis only on "@CA" atoms.
However, when you imaged the results in VMD, you used the full protein PDB file as coordinate input.
Thus, the vector lengths and projections in the .pev and .ppj files are being applied to the wrong atoms.
I see two options:
-Make a PDB file with only the CA atoms and use that as the input coordinates.
-Alternatively, carry out the PCA analysis on the entire protein.

I tried the first option and it works. I'm sure the second would as well.
As a note, you might want to confirm that the offset of 10 in the trajin command is not affecting your results.
It might give you lower resolution, but I don't know for sure.
I attached the CA-only PDB if you want to use it.

Good Luck,
Niel



_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber