btclusterer
Clusters purification bootstrap samples using the Markov Clustering
algorithm. Scans over a specific range of inflation coefficient
parameters.
This tool implements the only time-consuming step of bootstrap scores
network calculation. It can be used to parallelize the calculation
process by clustering only a small set of purification data samples for
each call. Use btfinisher
to merge the results and calculate the final bootstrap network and complexes.
Call the tool without any command line arguments to get detailed
information about the parameters for this program.
See also: Score network
generation
Example calls
These examples tie in with the bootstrap sample generation example from btsampler
. They assume an existing output directory btresults/
in the current working directory.
- use samples contained in
gavin_1000_samples
(-i
)
- cluster the samples 1 to 10 (
-s
), samples are 1-based
- use inflation coefficients 2.2 to 2.9 in steps of 0.1 (
-inf
)
- cut-off the generated scores network at 5.0 before clustering (
-c
)
- output to
btresults/
directory, files get the prefix bt_
(-prefix
)
./cmdline.sh btclusterer -i gavin_1000_samples -s 1-10 -inf 2.2,2.3,2.4,2.5,2.6,2.7,2.8,2.9 -c 5
-prefix btresults/bt_
- now cluster samples 11-20
./cmdline.sh btclusterer -i gavin_1000_samples -s 11-20 -inf 2.2,2.3,2.4,2.5,2.6,2.7,2.8,2.9 -c 5 -prefix btresults/bt_
and so on...
For each sample one network and one clustering file will be generated in the output directory.