Clustering of complexes
The SDA 6 distribution contains two algorithms for clustering docking results into representatives:

1. myClustering_complexes.py (in the auxi/scripts subdirectory)
2. clust (in the bin directory).

To run the first algorithm the following protocol can be used: # superimpose all structures on surface, only run on surface docking results!
auxi/scripts/DelRotZ_complexes.py complexes 2000
#clustering
myClustering_complexes.py wtrotz.complexes p2.pdb 2000
#generate representative pdb's
generateFortComplexesPdbs.py mycluster.complexes p2.pdb 3 X
The second clustering algorithm can be used in the following way: # superimpose all structures on surface, only run on surface docking results!
auxi/scripts/DelRotZ_complexes.py complexes 2000
#prepare cluster.dat file
bin/clust -cl -n 2000 -f55 wtrotz.complexes -p1 p1.pdb -p2 p2.pdb
#make pdb files for 7 mostly populated clusters
bin/clust -re -clcl 7 -f55 fort.55 -p1 p1.pdb -p2 p2.pdb
#prepare score file
./clust -sc -f55 wtrotz.complexes -p1 p1.pdb -p2 p2.pdb
#make energy analysis
./clust -an -p -clcl 7 -f55 wtrotz.complexes -p1 p1.pdb -p2 p2.pdb > cluster-out-info-2000
The second algorithm has more analysis and plotting options that are described by a seperate documentation in $SDA_DISTRIBUTION_DIR/auxi/clustering/README and $SDA_DISTRIBUTION_DIR/auxi/clustering/doc.

---

[Back to Index]