License: The MIT/Expat license
About: Run enrichment (permutation) test. It is efficient, does not attempt to place calls
in inaccessible regions, and correctly handles calls and regions of different sizes.
All coordinates are 1-based, inclusive, unless the file names have ".bed", ".bed.gz"
or ".bed.bgz" extension.
Usage: perm-test [OPTIONS]
Options:
-b, --background-regs FILE Background regions, expected not to be enriched: chr,beg,end
-c, --calls FILE Calls: chr,beg,end
-d, --debug-regions Print the spliced regions (and stop)
-f, --ref-fai FILE Chromosome lengths, given for example as fai index: chr,length
--no-bg-overlap Permuted variants must not overlap background regions
-o, --output FILE Place output in FILE
-m, --max-call-length INT Skip big calls [10e6]
-n, --niter NUM1[,NUM2] Number of iterations: total (NUM1) and per-batch (NUM2, reduces memory)
-s, --random-seed INT Random seed
-t, --target-regs FILE Target regions, expected to be enriched: chr,beg,end
Examples:
# Take a set of "background" regions (e.g. exome baits), "target" regions (e.g. a set of genes)
# and calculate how likely it is for a set of calls (e.g. CNVs) to overlap the target regions given
# the accessible background+target regions. The program ensures that overlapping and duplicate
# regions are handled correctly.
perm-test -b baits.txt -t genes.txt -c calls.txt -f ref.fai -n 1e9,1e8
# Same as above but this time count only permuted variants that overlap target regions but do not
# overlap any background region (e.g. CNVs that hit only a non-coding region but not a coding sequence).
# Note that target regions take precedence over background regions in case the sets overlap. If that is
# not desired, use `bedtools subtract -a noncoding.bed -b coding.bed` to clip and write regions from
# file A not present in B.
perm-test -b coding.bed -t noncoding.bed -c calls.txt -f ref.fai -n 1e9,1e8 --no-bg-overlap