Hello Pymol Users
Im reletalvely new to python, but have some questions regarding
pymols commands with regard to rot+trls matrices.
Ive used a small script to analyse domain motions, by first aligning
prot2 to prot1 domain1 and then make a copy of prot2 (which reset
object_matrix), align this copy to prot1 domain2 and get the object
matrix. This gives results as expected and verified with litterature.
However, I have a protein trajectory i would like to analyse, and
use the cmd.align(prot1+domselection, source_state=state,
target_state=1) twice as above for each domain. I have used the trick
with copying states which gives the correct matrix, but then I loose
the reference for the selection (from GUI) which I need for other
processing (And the copying also slows down the script). I might
process the selection with the iterate state before to get a resi list
which I can apply to the newly copied object, but the issue worries me
a bit.
I have tried to apply cmd.matrix_reset(prot1,state) after the first
alignment, but this gives wrong results. The matrix isnt unity(although
very close)
I have tried to extract the matrix after first alignment, and then
after second, and derive the alignment matrix using
matrixmultply(proM,inverse(preproM)), after conversion to arrays for
NumPy which gives similar results to cmd.matrix_reset. Both are not
consistent between runs and seem to depend on the objects matrix for
states before entering the script.
So if anyone can shed some light on why it works to copy the state
to new object after first alignment, but not cmd.matrix_reset of state
it would be great. There seem to be some other setting
(set_view/get_view???,origin for object???,???) which influence the
result of cmd.get_object_matrix, but I havent been able to figure out
what.
Best Regards
Esben Jannik Bjerrum
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
