if( $numfil != 0)then
foreach fnam (${DIR}/.pdb.?)
set fr=$fnam:r
set fnum=$fnam:e
mv $fnam $fr.0000$fnum
echo $fnam $fr.0000$fnum
end
endif
ls -1 ${DIR}/.pdb*>framelist
The second line of the script defines the PDB subdirectory and
should point at the one used in the previous step. The file created at
the end of the script (“framelist”) contains the list of all frames to be
used in subsequent clustering.
In the third step we launch kclust and create an output file
(“kclust.out”) with the results:
#!/bin/sh
rad=2.0
list=’framelist’
kclust -mode rmsd -centroid -cdist -heavy -lsqfit -radius $rad \
-maxerr 1 -iterate ${list}>kclust.out
ncl=‘grep Cluster kclust.out | wc -l‘
echo Found $ncl clusters
In the above, the name of the file with the list of frames
(“framelist”) should match the one used in the previous step. The
rad parameter is the radius in A ̊, defining the size of a cluster. If a
structure has RMSD distance from the centroid that is larger than
rad A ̊, it will not be included in the current cluster. By controlling
this parameter, we control the number of generated clusters.
In the fourth step we extract the centroids from the list of
clusters:
#!/bin/sh
awk -f extract_centroids.awk kclust.out | tee centroids.stat
Here we use an awk script (“extract_centroids.awk”, available
from http://ambermd.org/tutorials/basic/tutorial3/files/
extract_centroids.awk):
BEGIN{b0=2;}
{centind=index($1,"#Centroid");
c=$2;
getline;centind=index($0,"#Centroid");
FIL0 = sprintf("centroid%2.2d.member.dat",c)
while(centind != 1){
print $1,$3>FIL0 ;
Molecular Dynamics in Virtual Screening 159