Configuration file `experiment.json':
{
"Window Size" : 1000,
"Bin Size" : 10,
"Read Extension" : 0,
"Genome Path" : "hg19.genome",
"Track Path" : "experiment.track.wig",
}Generate a wiggle track file from *-seq reads
pshape -c experiment.json track experiment.bed
Extract experimental data from regions defined in 'experiment.foreground.000.bed' and estimate the foreground model
pshape -c experiment.json extract -a experiment.foreground.000.bed experiment.foreground.000.table
pshape -c experiment.json estimate --model="independent gamma" experiment.foreground.000.table experiment.foreground.000.json
Draw 10000 regions at random where any overlaps with the foreground are excluded. Afterwards, extract the experimental data and estimate the background model
pshape -c experiment.json draw -n 10000 --exclude=experiment.foreground.000.bed experiment.background.000.bed
pshape -c experiment.json extract experiment.background.000.bed experiment.background.000.table
pshape -c experiment.json estimate experiment.background.000.table experiment.background.000.json
Classify experimental data into foreground and background. The classification result is saved as a wiggle track file in 'experiment.classify.000.wig'. From this file, all regions classified as foreground are selected and saved in 'experiment.positive.000.table'
pshape -c experiment.json -t 30 classify experiment.foreground.000.json experiment.background.000.json experiment.classify.000.wig
pshape -c experiment.json positive experiment.classify.000.wig experiment.positive.000.table
Validate predictions by querying the user
pshape -c experiment.json validate --validated=experiment.foreground.000.bed experiment.positive.000.table experiment.foreground.001.bed