def alignmentTargets(genome_files, contig_files):
'''
generator object to produce filenames for
aligning contigs to known ncbi genomes
'''
parameters = []
for genome, contig in itertools.product(genome_files, contig_files):
outfile = os.path.join(
"alignment.dir",
P.snip(contig, ".contigs.fa") + "_vs_" +
P.snip(os.path.basename(genome), ".fna")) + ".delta"
additional_input = add_inputs(contig)
parameters.append([outfile, genome, contig])
return parameters
def filterContigsByCoverage(infiles, outfile):
'''
filter contigs by their average base coverage
'''
fcoverage = PARAMS["coverage_filter"]
contig_file = infiles[0]
dbh = sqlite3.connect(PARAMS["database"])
cc = dbh.cursor()
for infile in infiles[1:]:
print(contig_file, P.snip(os.path.basename(infile), ".load"))
def chimeraTargets(alignment_files, contig_files):
'''
generator object to produce filenames for
scoring chimericity
'''
parameters = []
for alignment, contig in itertools.product(genome_files, contig_files):
outfile = os.path.join("chimeras.dir",
P.snip(alignment, ".bam") + ".chimeras")
parameters.append([outfile, alignment, contig])
return parameters
def collectGenomeSizes(infile, outfile):
'''
output the genome sizes for each genome
'''
to_cluster = True
outf = open(outfile, "w")
outf.write("genome\tlength\n")
# assume single fasta entry
for fasta in FastaIterator.iterate(iotools.openFile(infile)):
name = P.snip(os.path.basename(infile), ".fna")
length = len(list(fasta.sequence))
outf.write("%s\t%s\n" % (name, str(length)))
outf.close()
def calculateFalsePositiveRate(infiles, outfile):
'''
calculate the false positive rate in taxonomic
abundances
'''
# connect to database
dbh = sqlite3.connect(PARAMS["database"])
cc = dbh.cursor()
true_file = infiles[0]
true_set = set()
estimate_set = set()
for estimate_file in infiles[1:]:
if os.path.basename(estimate_file)[
len("metaphlan_"):] == os.path.basename(true_file):
tablenames = [
P.toTable(os.path.basename(true_file)),
P.toTable(os.path.basename(estimate_file))
]
for species in cc.execute("""SELECT species_name FROM %s""" %
tablenames[0]).fetchall():
true_set.add(species[0])
for species in cc.execute(
"""SELECT taxon FROM %s WHERE taxon_level == 'species'""" %
tablenames[1]).fetchall():
if species[0].find("_unclassified") != -1: continue
estimate_set.add(species[0])
total_estimate = len(estimate_set)
total_true = len(true_set)
E.info("counting false positives and false negatives")
print(estimate_set.difference(true_set))
nfp = len(estimate_set.difference(true_set))
nfn = len(true_set.difference(estimate_set))
ntp = len(estimate_set.intersection(true_set))
E.info("writing results")
track = P.snip(os.path.basename(true_file), ".load")
outf = open(outfile, "w")
outf.write("track\ttp_rate\tfp_rate\tfn_rate\n")
outf.write("\t".join(
map(str, [
track,
float(ntp) / total_estimate,
float(nfp) / total_estimate,
float(nfn) / total_true
])) + "\n")
outf.close()
def alignContigsToReference(outfile, param1, param2):
'''
align the contigs to the reference genomes
using nucmer
'''
to_cluster = True
reffile, contigfile = param1, param2
pattern = P.snip(os.path.basename(outfile), ".delta")
statement = '''nucmer -p %(pattern)s %(reffile)s %(contigfile)s'''
P.run()
outf = os.path.basename(outfile)
statement = '''mv %(outf)s alignment.dir'''
P.run()
def buildAlignmentSizes(infiles, outfile):
'''
use bed files to sum the total number of bases
that are aligned to the genomes
'''
outf = open(outfile, "w")
outf.write("genome\tsize\n")
for infile in infiles:
genome = P.snip(os.path.basename(infile), ".bed.gz")
c = 0
inf = iotools.openFile(infile)
for bed in Bed.iterator(inf):
c += bed.end - bed.start
outf.write("%s\t%s\n" % (genome, str(c)))
outf.close()
def plotRelativeAbundanceCorrelations(infiles, outfile):
'''
plot the correlation between the estimated
relative abundance of species and the true
relative abundances - done on the shared set
'''
# connect to database
dbh = sqlite3.connect(PARAMS["database"])
cc = dbh.cursor()
true_file = infiles[0]
temp = P.getTempFile()
temp.write("true\testimate\n")
for estimate_file in infiles[1:]:
if os.path.basename(estimate_file)[
len("metaphlan_"):] == os.path.basename(true_file):
tablenames = [
P.toTable(os.path.basename(true_file)),
P.toTable(os.path.basename(estimate_file))
]
# get data
statement = """SELECT a.relab, b.rel_abundance
FROM %s as a, %s as b
WHERE b.taxon_level == "species"
AND a.species_name == b.taxon""" % (tablenames[0],
tablenames[1])
for data in cc.execute(statement).fetchall():
true, estimate = data[0], data[1]
temp.write("%f\t%f\n" % (true, estimate))
temp.close()
print(temp.name)
inf = temp.name
R('''data <- read.csv("%s", header = T, stringsAsFactors = F, sep = "\t")'''
% inf)
R('''png("%s")''' % outfile)
main_name = P.snip(outfile, ".png")
R('''data$estimate <- data$estimate/100''')
R('''plot(data$estimate, data$true, pch = 16, main = "%s", xlab = "estimated relative abundance", ylab = "observed relative abundance")'''
% main_name)
R('''text(0.05, y = 0.35, labels = paste("r = ", round(cor(data$estimate, data$true),2)), cex = 2)'''
)
R["dev.off"]()
os.unlink(inf)