def _get_contigs_to_plot(alignment_summ_gff, contigs):
"""
Returns a dict (string: ContigCoverage) that maps a contig ID to its coverage object.
:param alignment_summ_gff: (str) path to alignment_summ_gff
:param contigs: (list) top contigs from reference
"""
def _get_name(id_):
for c in contigs:
if c.id == id_:
return c.name
cov_map = {}
contig_ids = [c.id for c in contigs]
reader = GffReader(alignment_summ_gff)
for rec in reader:
if rec.seqid not in contig_ids:
log.info("Skipping seqid '{i}'.".format(i=rec.seqid))
continue
try:
contig_cov = cov_map[rec.seqid]
except KeyError:
contig_cov = ContigCoverage(rec.seqid, _get_name(rec.seqid))
cov_map[rec.seqid] = contig_cov
contig_cov.add_data(rec)
reader.close()
return cov_map
def _get_contigs_to_plot(alignment_summ_gff, contigs):
"""
Returns a dict (string: ContigCoverage) that maps a contig ID to its coverage object.
:param alignment_summ_gff: (str) path to alignment_summ_gff
:param contigs: (list) top contigs from reference
"""
def _get_name(id_):
for c in contigs:
if c.id == id_:
return c.name
cov_map = {}
contig_ids = [c.id for c in contigs]
reader = GffReader(alignment_summ_gff)
for rec in reader:
if rec.seqid not in contig_ids:
log.info(
"Unable to find gff '{i}' in alignment contig ids.".format(i=rec.seqid))
continue
try:
contig_cov = cov_map[rec.seqid]
except KeyError:
contig_cov = ContigCoverage(rec.seqid, _get_name(rec.seqid))
cov_map[rec.seqid] = contig_cov
contig_cov.add_data(rec)
reader.close()
return cov_map
def _get_contig_coverage(alignment_summ_gff, contigs):
"""
Modifies the passed contigs object to include coverage information.
:param alignment_summ_gff: (str) path to alignment_summ_gff
:param contigs: (dict) contig id -> ContigInfo object
"""
reader = GffReader(alignment_summ_gff)
for rec in reader:
# Some contigs don't have any coverage, but make it into the gff file
if rec.seqid in contigs:
contigs[rec.seqid].add_coverage_data(rec)
reader.close()
def _get_contig_coverage(alignment_summ_gff, contigs):
"""
Modifies the passed contigs object to include coverage information.
:param alignment_summ_gff: (str) path to alignment_summ_gff
:param contigs: (dict) contig id -> ContigInfo object
"""
reader = GffReader(alignment_summ_gff)
for rec in reader:
# Some contigs don't have any coverage, but make it into the gff file
if rec.seqid in contigs:
contigs[rec.seqid].add_coverage_data(rec)
reader.close()
def _extract_alignment_summ_data(aln_summ_gff, contigs):
"""
:param aln_summ_gff: (str) path to alignment_summary.gff
:param contigs: (list) top contigs from reference
:returns: 2 dictionaries containing data extracted from alignment_summary.gff
"""
def _get_name(id_):
for c in contigs:
if c.id == id_:
return c.name
contig_ids = [c.id for c in contigs]
ref_data = {}
var_map = {}
log.info("Reading GFF data from {f}".format(f=aln_summ_gff))
reader = GffReader(aln_summ_gff)
for rec in reader:
seqid = rec.seqid.split()[0]
if seqid not in contig_ids:
continue
# first data set
ref_data.setdefault(seqid, [0, 0, 0, 0])
ref_data[seqid][LENGTH] = max(rec.end, ref_data[seqid][LENGTH])
numGaps, lenGaps = rec.attributes["gaps"].split(",")
ref_data[seqid][GAPS] += int(lenGaps)
ref_data[seqid][COV] += float( rec.attributes["cov2"].split(",")[0] ) * \
(rec.end - rec.start + 1)
# second data set
contig_var = None
try:
contig_var = var_map[seqid]
except KeyError:
contig_var = ContigVariants(seqid, _get_name(seqid))
var_map[seqid] = contig_var
contig_var.add_data(rec)
reader.close()
return ref_data, var_map
def _append_variants_gff_data(ref_data, variants_gff):
"""
Adds data from variants gff to the ref_data dict
:param ref_data: (dict) dict of data pulled from alignment_summary.gff
:param variants_gff: (str) path to variants_gff
:type variants_gff: str
"""
reader = GffReader(variants_gff)
for record in reader:
err_len = record.end - record.start + 1
seqid = record.seqid.split()[0]
if seqid in ref_data:
ref_data[seqid][ERR] += err_len
else:
# the variants might not be present in the top 25 contigs,
# so we can just raise a warning in the log.
msg = "Unable to find {r} in {f}".format(r=seqid, f=variants_gff)
log.warn(msg)
reader.close()
def _append_variants_gff_data(ref_data, variants_gff):
"""
Adds data from variants gff to the ref_data dict
:param ref_data: (dict) dict of data pulled from alignment_summary.gff
:param variants_gff: (str) path to variants_gff
:type variants_gff: str
"""
reader = GffReader(variants_gff)
for record in reader:
err_len = record.end - record.start + 1
seqid = record.seqid.split()[0]
if seqid in ref_data:
ref_data[seqid][ERR] += err_len
else:
# the variants might not be present in the top 25 contigs,
# so we can just raise a warning in the log.
msg = "Unable to find {r} in {f}".format(
r=seqid, f=variants_gff)
log.warn(msg)
reader.close()
def main():
headers = [
("source", "GenomicConsensus %s" % __VERSION__),
("pacbio-alignment-summary-version", "0.6"),
("source-commandline", " ".join(sys.argv)),
]
desc = "Augment the alignment_summary.gff file with consensus and variants information."
parser = argparse.ArgumentParser(description=desc)
parser.add_argument("--variantsGff",
type=str,
help="Input variants.gff or variants.gff.gz filename",
required=True)
parser.add_argument("--output",
"-o",
type=str,
help="Output alignment_summary.gff filename")
parser.add_argument("inputAlignmentSummaryGff",
type=str,
help="Input alignment_summary.gff filename")
options = parser.parse_args()
inputVariantsGff = GffReader(options.variantsGff)
inputAlignmentSummaryGff = GffReader(options.inputAlignmentSummaryGff)
summaries = {}
for gffRecord in inputAlignmentSummaryGff:
region = Region(gffRecord.seqid, gffRecord.start, gffRecord.end)
summaries[region] = {"ins": 0, "del": 0, "sub": 0, "cQv": (0, 0, 0)}
inputAlignmentSummaryGff.close()
counterNames = {
"insertion": "ins",
"deletion": "del",
"substitution": "sub"
}
for variantGffRecord in inputVariantsGff:
for region in summaries:
summary = summaries[region]
if (region.seqid == variantGffRecord.seqid
and region.start <= variantGffRecord.start <= region.end):
counterName = counterNames[variantGffRecord.type]
variantLength = max(len(variantGffRecord.reference),
len(variantGffRecord.variantSeq))
summary[counterName] += variantLength
# TODO: base consensusQV on effective coverage
summary["cQv"] = (20, 20, 20)
inputAlignmentSummaryGff = open(options.inputAlignmentSummaryGff)
outputAlignmentSummaryGff = open(options.output, "w")
inHeader = True
for line in inputAlignmentSummaryGff:
line = line.rstrip()
# Pass any metadata line straight through
if line[0] == "#":
print >> outputAlignmentSummaryGff, line.strip()
continue
if inHeader:
# We are at the end of the header -- write the tool-specific headers
for k, v in headers:
print >> outputAlignmentSummaryGff, ("##%s %s" % (k, v))
inHeader = False
# Parse the line
rec = Gff3Record.fromString(line)
if rec.type == "region":
summary = summaries[(rec.seqid, rec.start, rec.end)]
if "cQv" in summary:
cQvTuple = summary["cQv"]
line += ";%s=%s" % ("cQv", ",".join(
str(int(f)) for f in cQvTuple))
for counterName in counterNames.values():
if counterName in summary:
line += ";%s=%d" % (counterName, summary[counterName])
print >> outputAlignmentSummaryGff, line
