I am running structure prediction of a protein whose tertiary structure is unknown. I use the standard abinitio protocol explained in the Rosetta tutorial(s). I have currently generated 3000 decoys, which I tried to cluster in order to have an idea of which structures were mostly populated.
Using the following command for clustering:
cluster.default.linuxgccrelease -in:file:fullatom -cluster:radius 3 -nooutput -out:file:silent ./cluster.out -in:file:s *.pdb
I obtain an output file with only 173 entries/structures. I was expecting 3000 structures, exactly like the number of decoys, but renamed as c.i.j.pdb as per Rosetta convention. Why is this not happening? Is there something wrong in the command line above?
What I would like to achieve is to cluster all my decoys (they will be more than 3000) and obtain (i) the number of clusters and (ii) their populations. The hope is that the most populated cluster is also the folded structure of my protein (we have some experimental information on the overall shape of this protein).
Any suggestions is highly appreciated.