def bamparse((strain, target, bamfile, targetname)):
"""Parses bam files using pysam stats"""
import pysamstats
parsedict = defaultdict(bamPysamStats.make_dict)
# Use the stat_baseq_ext (extended base quality statistics) function of pysam stats to return records parsed
# from sorted bam files
for rec in pysamstats.stat_baseq_ext(alignmentfile=bamfile, fafile=target):
# Values of interest can be retrieved using the appropriate keys
# Simple filtering statement: if the number of matches at a particular position in the reference sequence is
# greater than the number of mismatches, and the total depth is 5 or more, add the position of the results
if rec['matches'] > rec['mismatches'] and rec['reads_all'] > 4:
# Populate the dictionary with the appropriate values
parsedict[strain][target][rec['chrom']][float(rec['pos'])][rec['reads_all']] = rec['rms_baseq']
bamPysamStats.dotter()
# dotter()
return parsedict
def bamparse((strain, target, bamfile, targetname)):
"""Parses bam files using pysam stats"""
import pysamstats
parsedict = defaultdict(bamPysamStats.make_dict)
# Use the stat_baseq_ext (extended base quality statistics) function of pysam stats to return records parsed
# from sorted bam files
for rec in pysamstats.stat_baseq_ext(alignmentfile=bamfile, fafile=target):
# Values of interest can be retrieved using the appropriate keys
# Simple filtering statement: if the number of matches at a particular position in the reference sequence is
# greater than the number of mismatches, and the total depth is 5 or more, add the position of the results
if rec['matches'] > rec['mismatches'] and rec['reads_all'] > 4:
# Populate the dictionary with the appropriate values
parsedict[strain][target][rec['chrom']][float(
rec['pos'])][rec['reads_all']] = rec['rms_baseq']
bamPysamStats.dotter()
# dotter()
return parsedict
def parsearmi(self):
"""
Multiprocessing for parsing bam files
:param seqdict: dictionary containing import path and name information of files and folders
:param analysistype: string of the analysis type
"""
from glob import glob
# Initialise dictionary, argument list, and Pool
loadedresultsdict = defaultdict(bamPysamStats.make_dict)
bamparseprocessesargs = []
bamparseprocessespool = Pool()
# Iterate through the strains
for strain in self.seqdict:
# Store the identityCutoff in seqDict
self.seqdict[strain]["cutoff"][
self.analysistype] = self.identitycutoff
# Retrieve bait type and determine the directory of the fastq files from seqdict
baittype = self.seqdict[strain]["bait"]["fastqFiles"].keys()[0]
fastqdir = os.path.split(
self.seqdict[strain]["bait"]["fastqFiles"][baittype][0])[0]
# Set the name of the JSON file to store the results
jsonprofile = "%s/%s_matchDict.json" % (fastqdir,
self.analysistype)
# If the JSON file hasn't been created, parse the bam files
if not os.path.isfile(jsonprofile):
bamfile = glob(
'{}/ARMIConcatenated/*_sorted.bam'.format(fastqdir))[0]
for target in self.seqdict[strain]["targets"][
self.analysistype]:
# Get the name of the target from the target variable
targetname = os.path.basename(target).split(".")[0]
# Set the input/output dir
# outdir = "%s/%s" % (fastqdir, targetname)
# Make a list of the sorted bam files
# bamfile = glob("%s/*_sorted.bam" % outdir)[0]
# catfile = self.seqdict[strain]["concatenatedTargets"][self.analysistype]
# print "!", catfile
# Append a tuple of the required arguments to the argument list
bamparseprocessesargs.append(
(strain, target, bamfile, targetname))
# print strain, target, bamfile
# If the JSON file exists, read the results from it rather than performing the parsing again
else:
# Open the JSON file
with open(jsonprofile, "rb") as jsonreport:
# Load the results from the JSON file into a dictionary
loadedresultsdict[strain].update(json.load(jsonreport))
bamPysamStats.dotter()
# Run the multiprocessed bam parsing
parselist = bamparseprocessespool.map(bamparse, bamparseprocessesargs)
# Change the returned list of dictionaries into a nested dictionary
self.parseddict = bamPysamStats.filler(parselist)
# Load the length of the targets using the .fai files generated in the bamParse function
self.seqdict = bamPysamStats.targetlength(self.seqdict,
self.analysistype)
# Iterate through the strains in order to write the results to a JSON file
for strain in self.seqdict:
# Get the bait type
baittype = self.seqdict[strain]["bait"]["fastqFiles"].keys()[0]
fastqdir = os.path.split(
self.seqdict[strain]["bait"]["fastqFiles"][baittype][0])[0]
# Define the JSON profile file
jsonprofile = "%s/%s_matchDict.json" % (fastqdir,
self.analysistype)
# If the file doesn't exist, create it, and fill it with results
if not os.path.isfile(jsonprofile):
jsonreport = open(jsonprofile, "wb")
output = json.dumps(self.parseddict[strain],
sort_keys=True,
indent=4,
separators=(',', ': '))
jsonreport.write(output)
jsonreport.close()
self.armiparser()
def parsearmi(self):
"""
Multiprocessing for parsing bam files
:param seqdict: dictionary containing import path and name information of files and folders
:param analysistype: string of the analysis type
"""
from glob import glob
# Initialise dictionary, argument list, and Pool
loadedresultsdict = defaultdict(bamPysamStats.make_dict)
bamparseprocessesargs = []
bamparseprocessespool = Pool()
# Iterate through the strains
for strain in self.seqdict:
# Store the identityCutoff in seqDict
self.seqdict[strain]["cutoff"][self.analysistype] = self.identitycutoff
# Retrieve bait type and determine the directory of the fastq files from seqdict
baittype = self.seqdict[strain]["bait"]["fastqFiles"].keys()[0]
fastqdir = os.path.split(self.seqdict[strain]["bait"]["fastqFiles"][baittype][0])[0]
# Set the name of the JSON file to store the results
jsonprofile = "%s/%s_matchDict.json" % (fastqdir, self.analysistype)
# If the JSON file hasn't been created, parse the bam files
if not os.path.isfile(jsonprofile):
bamfile = glob('{}/ARMIConcatenated/*_sorted.bam'.format(fastqdir))[0]
for target in self.seqdict[strain]["targets"][self.analysistype]:
# Get the name of the target from the target variable
targetname = os.path.basename(target).split(".")[0]
# Set the input/output dir
# outdir = "%s/%s" % (fastqdir, targetname)
# Make a list of the sorted bam files
# bamfile = glob("%s/*_sorted.bam" % outdir)[0]
# catfile = self.seqdict[strain]["concatenatedTargets"][self.analysistype]
# print "!", catfile
# Append a tuple of the required arguments to the argument list
bamparseprocessesargs.append((strain, target, bamfile, targetname))
# print strain, target, bamfile
# If the JSON file exists, read the results from it rather than performing the parsing again
else:
# Open the JSON file
with open(jsonprofile, "rb") as jsonreport:
# Load the results from the JSON file into a dictionary
loadedresultsdict[strain].update(json.load(jsonreport))
bamPysamStats.dotter()
# Run the multiprocessed bam parsing
parselist = bamparseprocessespool.map(bamparse, bamparseprocessesargs)
# Change the returned list of dictionaries into a nested dictionary
self.parseddict = bamPysamStats.filler(parselist)
# Load the length of the targets using the .fai files generated in the bamParse function
self.seqdict = bamPysamStats.targetlength(self.seqdict, self.analysistype)
# Iterate through the strains in order to write the results to a JSON file
for strain in self.seqdict:
# Get the bait type
baittype = self.seqdict[strain]["bait"]["fastqFiles"].keys()[0]
fastqdir = os.path.split(self.seqdict[strain]["bait"]["fastqFiles"][baittype][0])[0]
# Define the JSON profile file
jsonprofile = "%s/%s_matchDict.json" % (fastqdir, self.analysistype)
# If the file doesn't exist, create it, and fill it with results
if not os.path.isfile(jsonprofile):
jsonreport = open(jsonprofile, "wb")
output = json.dumps(self.parseddict[strain], sort_keys=True, indent=4, separators=(',', ': '))
jsonreport.write(output)
jsonreport.close()
self.armiparser()
