Cluster a segment (domain) of a protein by superposing different part (domain) of the protein using their three-dimensional coordinates.
- mod9.21
- python2 or python3
- Scikit-learn
- Required distances were computed using mod9.21.
- Subsequent clustering was performed using sklearn.cluster.DBSCAN (Density-based spatial clustering of applications with noise ).
- Precomputed distance matrix using mod9.21 was inputted in DBSCAN.
Clusters computed using DBSCAN were further matched with Gromacs for testing purpose. Gromacs implements single linkage clustering algorithm by default. Clusters produced by both methods are statistically indistinguishable.
$ python ClusterDistMatrix.py -i ./unittest/rmsd.txt.tgz