def extractReads(self, bams=[]):
"""Extract reads from sam files and write to file"""
# load data
self.BM = binManager.BinManager(dbFileName=self.dbFileName) # bins
self.BM.loadBins(makeBins=True,silent=False,bids=self.bids)
self.PM = self.BM.PM
print "Extracting reads"
# work out a set of targets to pass to the parser
targets = {}
bids = self.BM.getBids()
for bid in bids:
bin = self.BM.getBin(bid)
for row_index in bin.rowIndices:
targets[self.PM.contigNames[row_index]] = bid
# get something to parse the bams with
bam_parser = BTBP()
bam_parser.extractReads(bams,
'',
targets,
combineBams=False,
headersOnly = True,
dontTrustSamFlags=False,
folder=self.outDir,
verbose=True
)
def parse(self, bamFiles, stoitColNames, covTable, contigNames, cid2Indicies):
"""Parse multiple bam files and store the results in the main DB
table: a table in an open h5 file like "CID,COV_1,...,COV_n,length"
stoitColNames: names of the COV_x columns
"""
print "Importing BAM files"
from bamtyper.utilities import BamParser as BTBP
BP = BTBP()
(links, ref_lengths, coverages) = BP.getLinks(bamFiles, full=False, verbose=True, doCoverage=True, minJoin=5)
# go through all the contigs sorted by name and write to the DB
try:
for cid in sorted(contigNames.keys()):
# make a new row
cov_row = covTable.row
# punch in the data
for i in range(len(stoitColNames)):
try:
cov = coverages[i][cid]
except KeyError:
# may be no coverage for this contig
cov = 0.0
cov_row[stoitColNames[i]] = cov
cov_row.append()
covTable.flush()
except:
print "Error saving results to DB"
raise
# transform the links into something a little easier to parse later
rowwise_links = []
for cid in links:
for link in links[cid]:
try:
rowwise_links.append([cid2Indicies[cid], # contig 1
cid2Indicies[link[0]], # contig 2
int(link[1]), # numReads
int(link[2]), # linkType
int(link[3]) # gap
])
except KeyError:
pass
return rowwise_links
