README

BEDOPS enabled Hotspot. The software is a fast implementation of the hotspot algorithm for identifying regions of enrichment in next generation sequencing assays.

Requirements

• BEDOPS >2.4 (http://bedops.readthedocs.org/)
• Python 2.7 (w/ numpy)
• random-lines (from https://bitbucket.org/jvierstra/bio-tools)

Running

The software is modular -- various tasks will utlize somewhat different scripts.

Example on a single core:

# simulate data for FDR calculation
bash /home/jvierstra/proj/code/hotspot/scripts/simulate.cutcounts.sh --tmpdir=/tmp --starch-output fragments.starch ${GENOME_MAPPABILITY_FILE} cutcounts.simulated.starch

# Call hotspots on observed data
bash ${HOTSPOT_DIR}/scripts/hotspot.run.sh --tmpdir=/tmp cutcounts.starch ${GENOME_MAPPABILITY_FILE} hotspots.unthresholded.bed

# Call hotspots on simulated data
bash ${HOTSPOT_DIR}/scripts/hotspot.run.sh --tmpdir=/tmp cutcounts.simulated.starch ${GENOME_MAPPABILITY_FILE} hotspots.simulated.bed

# Compute Q-values 
python ${HOTSPOT_DIR}/scripts/compute_q_values.py hotspots.simulated.bed hotspots.unthresholded.bed > hotspots.fdr.bed

# Select threshold and merge
bash ${HOTSPOT_DIR}/scripts/hotspot.thresh-merge.sh --q-thresh=0.01 hotspots.fdr.bed hotspots.fdr.0.01.bed 

Example to run chromosomes on separate processes (e.g., cluster):

for i in $(seq 1 22); do
	chrom="chr$i"
	bash ${HOTSPOT_DIR}/scripts/hotspot.run.sh --tmpdir=/tmp --contig=$chrom cutcounts.starch ${GENOME_MAPPABILITY_FILE} hotspots.${chrom}.unthresholded.bed
done

• The "--contig=" option works on both the hotspot and simulate tags scripts.

TODO

• Badspot implementation
• Change downsample script to downsample from fragments, not cutcounts