def main():
setupLogging(fi="psvpTools.log", debug=False)
logging.info("^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^")
parseArgs(sys.argv[1:])
p = Psvp(inputFile, name, callers, annotationColumns)
outpath = outputDirectory + "/" + p.name
#All sites
#sizeDistribution
pp(sizeDistribution(p, sizeBins=SIZE_BINS), outpath + ".all.sizeDist.txt")
#callerDistribution
pp(callerDistribution(p), outpath + ".all.callDist.txt")
#sampleDistribution
pp(sampleDistribution(p), outpath + ".all.sampDist.txt")
#'Well-supported' sites - i.e. multicaller support & multisample support
p = findCallerSupport(p, minCallers=2)
p.clearEmptySites(mask=False)
p = findSampleSupport(p, minSamples=2, mask=False)
#sizeDistribution
pp(sizeDistribution(p, sizeBins=SIZE_BINS), outpath + ".sup.sizeDist.txt")
#callerDistribution
pp(callerDistribution(p), outpath + ".sup.callDist.txt")
#sampleDistribution
pp(sampleDistribution(p), outpath + ".sup.sampDist.txt")
logging.info("$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$")
def main():
setupLogging(fi="psvpTools.log")
logging.info("^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^")
p = Psvp(inputFile=INPUT_FILE, name=NAME, callers=CALLERS, annotationColumns=ANNOTATION_COLUMNS)
outpath = OUT_DIR + "/" + NAME
with open(outpath + "totals.txt", "w") as out:
out.write("") # need to clear this file to later append some output
# All sites
# Here, we look at all sites in our data, which consists of a .psvp for
# each chromosome concatenated together into one file, 'data.psvp'.
# psvpTools's class Psvp will keep the first header it comes across and
# discard the rest - important for if you are trying to output a
# filtered psvp for use with other tools, possibly in other formats.
with open(outpath + "totals.txt", "a") as out:
# Let's get a count for every category and put it in this file, starting
# with 'All sites'.
out.write("All:\t")
out.write(str(countSites(p)) + "\n")
#'prettyPrint' a size, caller, and sample distribution
pp(sizeDistribution(p, sizeBins=SIZE_BINS), outpath + ".all.sizeDist.txt")
pp(callerDistribution(p), outpath + ".all.callDist.txt")
pp(sampleDistribution(p), outpath + ".all.sampDist.txt")
#'Well-supported' sites - i.e. multicaller support & multisample support
# Now that we have some information on all sites, we want to look more
# closely at sites we're more confident aren't errors. So, let's:
# 1. Filter out all entries that aren't supported by at least two of our
# programs
# 2. Remove any sites which may not contain any nonzero entries now that
# we've modified them. We set mask=False to delete empty sites rather
# than mask them, since we won't be using them later.
# 3. Filter out sites that don't have at least 2 samples with entries.
p = findCallerSupport(p, minCallers=2)
p.clearEmptySites(mask=False)
p = findSampleSupport(p, minSamples=2, mask=False)
with open(outpath + "totals.txt", "a") as out:
out.write("Supported:\t")
out.write(str(countSites(p)) + "\n")
pp(sizeDistribution(p, sizeBins=SIZE_BINS), outpath + ".sup.sizeDist.txt")
pp(callerDistribution(p), outpath + ".sup.callDist.txt")
pp(sampleDistribution(p), outpath + ".sup.sampDist.txt")
#'Well-supported' sites NOT in the reference (sample column 73)
# Finally, let's focus on sites which don't show up as variants in the
# reference rhesus, which was included in the data so that we might
# filter out more false calls from our programs. The reference rhesus was
# the 74th sample column, or 73 columns away from the start of our
# the sample columns, so we defined annotationColumns as [73] when
# we created our Psvp. The reference data was left out of all previous
# psvpStat function calls because we had marked that column as an
# annotation.
# Now, we filter out sites which have nonzero data in the reference sample.
# We set reverse=True because we want sites that DO NOT have annotations.
p = findAnnotated(p, reverse=True, mask=False)
with open(outpath + "totals.txt", "a") as out:
out.write("Supported and not in reference:\t")
out.write(str(countSites(p)) + "\n")
pp(sizeDistribution(p, sizeBins=SIZE_BINS), outpath + ".snr.sizeDist.txt")
pp(callerDistribution(p), outpath + ".snr.callDist.txt")
pp(sampleDistribution(p), outpath + ".snr.sampDist.txt")
logging.info("$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$")
