% cd crd
% ln -s ../1fsc_dimens.stream ../1fsc.psf .
% setenv toppar ".."
% charmm pdbfile=1fsc begin=1 end=8 task=p2 \
< ../traj.inp > p2.out
This run will produce a file of the form:
1fsc_e1-8_p2_##.p2
, where ## is the residue number, one file for each residue
with a distinct NH backbone vector (not prolines or N-terminus). In this
case there are 55 files.
Here they are plotted 15 at a time (I used pastem
to make 4 files):
Order parameters are the assymptotic values of these series and
one can observe that most have
converged to around .8 to .9. On the other hand, there are some that
are either clearly not converged
or are questionable. Regardless of whether they are converged, we can
usually get an upper limit on the
value of order parameters by averaging the last portion of the curve
as the following awk script will do:
p2_to_orders <# pts to avg> <#pts to shift back for diff> <# residues> <filename root>
where,
<# pts to avg> is the number of final points in the p2 curve to average over
as an estimate of the order parameter.
<#pts to shift back for diff> is how many points to shift back toward the beginning of the series
to get a second estimate of the order parameter to compare to the first.
the difference between this second estimate and the first (better) estimate
is printed as the third column in the output and is an estimate of how
well the series has converged.
<# residues> total number of residues
<filename root> the filename up to the residue number (e.g. 1fsc_e1-8_p2_ )% p2_to_orders 75 75 61 1fsc_e1-8_p2_ > Orders% xmgrace -nxy Orders