MDFF for huge trajectories

From: Samuel Flores (
Date: Wed May 16 2012 - 14:31:53 CDT

Hi Guys,

I have a trajectory of the ribosome with 16S, 23S, two tRNAs, L5, and L1. I have a few hundred frames of this, in a .pdb formatted trajectory. I also have a few density maps that this trajectory needs to be compared against (trjectory should gradually fit into map). I am using this TCL script:

package require mdff;
for {set i 3} {$i < 102} {incr i } {
set sel0 [atomselect 0 " chain V or chain Y or chain A or chain a or chain c or chain g " frame i ];
mdff ccc $sel0 -i pre2.situs -res 5

I have two questions:

1. Isn't there a more efficient way to do this? It looks to me like it's loading the density map at every single frame. Surely the map could be loaded once and then referred to at each step $i?
2. How can I print out the cross-correlation coefficient? if I put the commands in through the TK console I get a printout, but if I run in batchmode (-dispdev text -e .. etc) then I try:

package require mdff;
puts "howdy!";
# load starting from frame *** (default 0):
mol new temp.pdb first 0 step 1 ;
for {set i 3} {$i <= 102} {incr i } {set sel0 [atomselect 0 " chain V or chain Y or chain A or chain a or chain c or chain g " frame i ];
set $crosscoefficient [mdff ccc $sel0 -i pre2.situs -res 5] ;
puts "Cross Correlation Coefficient (CCC) = ", $crosscoefficient;

but while this appears to still be doing the ccc calculation, it doesn't output the answer as expected. That second puts command appears to be doing nothing, though the first puts command works as expected.



This archive was generated by hypermail 2.1.6 : Tue Dec 31 2013 - 23:21:59 CST