def calculateCodingDensity(self, outDir, scaffolds, genomeSize):
"""Calculate coding density of putative genome bin."""
gffFile = os.path.join(outDir, DefaultValues.PRODIGAL_GFF)
if os.path.exists(gffFile):
prodigalParserGFF = ProdigalGeneFeatureParser(gffFile)
aaFile = os.path.join(outDir, DefaultValues.PRODIGAL_AA) # use AA file as nucleotide file is optional
aaGenes = readFasta(aaFile)
codingBasePairs = 0 # self.__calculateCodingBases(aaGenes)
for scaffold_id in scaffolds.keys():
codingBasePairs += prodigalParserGFF.codingBases(scaffold_id)
return float(codingBasePairs) / genomeSize, prodigalParserGFF.translationTable, len(aaGenes)
else:
# there is no gene feature file (perhaps the user specified pre-calculated genes)
# so calculating the coding density is not possible
return -1, -1, -1
def calculateCodingDensity(self, outDir, scaffolds, genomeSize):
"""Calculate coding density of putative genome bin."""
gffFile = os.path.join(outDir, DefaultValues.PRODIGAL_GFF)
if os.path.exists(gffFile):
prodigalParserGFF = ProdigalGeneFeatureParser(gffFile)
aaFile = os.path.join(
outDir, DefaultValues.PRODIGAL_AA
) # use AA file as nucleotide file is optional
aaGenes = readFasta(aaFile)
codingBasePairs = 0 # self.__calculateCodingBases(aaGenes)
for scaffold_id in scaffolds.keys():
codingBasePairs += prodigalParserGFF.codingBases(scaffold_id)
return float(
codingBasePairs
) / genomeSize, prodigalParserGFF.translationTable, len(aaGenes)
else:
# there is no gene feature file (perhaps the user specified pre-calculated genes)
# so calculating the coding density is not possible
return -1, -1, -1
def plotOnAxes(self, fastaFile, distributionsToPlot, axesHist, axesDeltaCD):
# parse Prodigal output
gffFile = os.path.join(self.options.out_folder, 'bins', binIdFromFilename(fastaFile), DefaultValues.PRODIGAL_GFF)
if not os.path.exists(gffFile):
print 'Missing gene feature file (%s). This plot if not compatible with the --genes option.' % DefaultValues.PRODIGAL_GFF
sys.exit()
prodigalParser = ProdigalGeneFeatureParser(gffFile)
# Read reference distributions from file
dist = readDistribution('cd_dist')
# get coding density for windows
seqs = readFasta(fastaFile)
data = []
seqLens = []
for seqId, seq in seqs.iteritems():
start = 0
end = self.options.cd_window_size
seqLen = len(seq)
seqLens.append(seqLen)
while(end < seqLen):
codingBases = prodigalParser.codingBases(seqId, start, end)
a, c, g, t = baseCount(seq[start:end])
data.append(float(codingBases) / (a + c + g + t))
start = end
end += self.options.cd_window_size
if len(data) == 0:
axesHist.set_xlabel('[Error] No seqs >= %d, the specified window size' % self.options.cd_window_size)
return
# Histogram plot
bins = [0.0]
binWidth = self.options.cd_bin_width
binEnd = binWidth
while binEnd <= 1.0:
bins.append(binEnd)
binEnd += binWidth
axesHist.hist(data, bins=bins, normed=True, color=(0.5, 0.5, 0.5))
axesHist.set_xlabel('% coding density')
axesHist.set_ylabel('% windows (' + str(self.options.cd_window_size) + ' bp)')
# Prettify plot
for a in axesHist.yaxis.majorTicks:
a.tick1On = True
a.tick2On = False
for a in axesHist.xaxis.majorTicks:
a.tick1On = True
a.tick2On = False
for line in axesHist.yaxis.get_ticklines():
line.set_color(self.axesColour)
for line in axesHist.xaxis.get_ticklines():
line.set_color(self.axesColour)
for loc, spine in axesHist.spines.iteritems():
if loc in ['right', 'top']:
spine.set_color('none')
else:
spine.set_color(self.axesColour)
# get CD bin statistics
binTools = BinTools()
meanCD, deltaCDs, _ = binTools.codingDensityDist(seqs, prodigalParser)
# Delta-CD vs sequence length plot
axesDeltaCD.scatter(deltaCDs, seqLens, c=abs(deltaCDs), s=10, lw=0.5, cmap=pylab.cm.Greys)
axesDeltaCD.set_xlabel(r'$\Delta$ CD (mean coding density = %.1f%%)' % (meanCD * 100))
axesDeltaCD.set_ylabel('Sequence length (kbp)')
_, yMaxSeqs = axesDeltaCD.get_ylim()
xMinSeqs, xMaxSeqs = axesDeltaCD.get_xlim()
# plot reference distributions
for distToPlot in distributionsToPlot:
closestCD = findNearest(np.array(dist.keys()), meanCD)
# find closest distribution values
sampleSeqLen = dist[closestCD].keys()[0]
d = dist[closestCD][sampleSeqLen]
cdLowerBoundKey = findNearest(d.keys(), (100 - distToPlot) / 2.0)
cdUpperBoundKey = findNearest(d.keys(), (100 + distToPlot) / 2.0)
xL = []
xU = []
y = []
for windowSize in dist[closestCD]:
xL.append(dist[closestCD][windowSize][cdLowerBoundKey])
xU.append(dist[closestCD][windowSize][cdUpperBoundKey])
y.append(windowSize)
# sort by y-values
sortIndexY = np.argsort(y)
xL = np.array(xL)[sortIndexY]
xU = np.array(xU)[sortIndexY]
y = np.array(y)[sortIndexY]
axesDeltaCD.plot(xL, y, 'r--', lw=0.5, zorder=0)
axesDeltaCD.plot(xU, y, 'r--', lw=0.5, zorder=0)
# ensure y-axis include zero and covers all sequences
axesDeltaCD.set_ylim([0, yMaxSeqs])
# ensure x-axis is set appropriately for sequences
axesDeltaCD.set_xlim([xMinSeqs, xMaxSeqs])
# draw vertical line at x=0
axesDeltaCD.vlines(0, 0, yMaxSeqs, linestyle='dashed', color=self.axesColour, zorder=0)
# Change sequence lengths from bp to kbp
yticks = axesDeltaCD.get_yticks()
kbpLabels = []
for seqLen in yticks:
label = '%.1f' % (float(seqLen) / 1000)
label = label.replace('.0', '') # remove trailing zero
kbpLabels.append(label)
axesDeltaCD.set_yticklabels(kbpLabels)
# Prettify plot
for a in axesDeltaCD.yaxis.majorTicks:
a.tick1On = True
a.tick2On = False
for a in axesDeltaCD.xaxis.majorTicks:
a.tick1On = True
a.tick2On = False
for line in axesDeltaCD.yaxis.get_ticklines():
line.set_color(self.axesColour)
for line in axesDeltaCD.xaxis.get_ticklines():
line.set_color(self.axesColour)
for loc, spine in axesDeltaCD.spines.iteritems():
if loc in ['right', 'top']:
spine.set_color('none')
else:
spine.set_color(self.axesColour)
def __calculateResults(self, windowSizes, numWindows, genomeDir, queueIn, queueOut):
while True:
genomeId = queueIn.get(block=True, timeout=None)
if genomeId == None:
break
seqs = readGenomicSeqsFromFasta(os.path.join(genomeDir, genomeId, genomeId + '.fna'))
# for simplicity, create a single scaffold from all sequences
genomeFile = os.path.join('./deltaCD/genomes', genomeId + '.single_scaffold.fna')
genomeScaffold = 'NNNNNNNNNN'.join(list(seqs.values())).upper()
fout = open(genomeFile, 'w')
fout.write('>' + genomeId + '\n')
fout.write(genomeScaffold)
fout.close()
# run prodigal on genome
ntFile = os.path.join('./deltaCD/prodigal', genomeId + '.genes.fna')
gffFile = os.path.join('./deltaCD/prodigal', genomeId + '.gff')
cmd = ('prodigal -q -c -m -f gff -d %s -i %s > %s' % (ntFile, genomeFile, gffFile))
os.system(cmd)
# calculate mean coding density of genome
numericScaffold = self.__createNumericScaffold(genomeScaffold)
prodigalParser = ProdigalGeneFeatureParser(gffFile)
codingBases = prodigalParser.codingBases(genomeId)
counts = np.bincount(numericScaffold)
totalBases = counts[0]
meanCD = float(codingBases) / totalBases
fout = open('./deltaCD/' + genomeId + '.tsv', 'w')
fout.write('# Mean CD = ' + str(meanCD) + '\n')
# calculate coding density distribution for different window sizes
for windowSize in windowSizes:
endWindowPos = len(genomeScaffold) - windowSize
if endWindowPos <= 0:
# This might occur for the largest window sizes and smallest genomes
break
deltaCDs = []
while len(deltaCDs) != numWindows:
# pick random window
startWindow = randint(0, endWindowPos)
# calculate coding density
codingBases = prodigalParser.codingBases(genomeId, startWindow, startWindow+windowSize)
counts = np.bincount(numericScaffold[startWindow:(startWindow+windowSize)])
totalBases = counts[0]
if totalBases != windowSize:
# there are N's in the window so skip it
continue
cdPer = float(codingBases) / totalBases
deltaCDs.append(cdPer - meanCD)
fout.write('Windows Size = ' + str(windowSize) + '\n')
fout.write(','.join(map(str, deltaCDs)) + '\n')
fout.close()
queueOut.put(genomeId)
def plotOnAxes(self, fastaFile, distributionsToPlot, axesHist, axesDeltaCD):
# parse Prodigal output
gffFile = os.path.join(self.options.results_dir, 'bins', binIdFromFilename(fastaFile), DefaultValues.PRODIGAL_GFF)
if not os.path.exists(gffFile):
self.logger.error('Missing gene feature file (%s). This plot if not compatible with the --genes option.' % DefaultValues.PRODIGAL_GFF)
sys.exit()
prodigalParser = ProdigalGeneFeatureParser(gffFile)
# Read reference distributions from file
dist = readDistribution('cd_dist')
# get coding density for windows
seqs = readFasta(fastaFile)
data = []
seqLens = []
for seqId, seq in seqs.iteritems():
start = 0
end = self.options.cd_window_size
seqLen = len(seq)
seqLens.append(seqLen)
while(end < seqLen):
codingBases = prodigalParser.codingBases(seqId, start, end)
a, c, g, t = baseCount(seq[start:end])
data.append(float(codingBases) / (a + c + g + t))
start = end
end += self.options.cd_window_size
if len(data) == 0:
axesHist.set_xlabel('[Error] No seqs >= %d, the specified window size' % self.options.cd_window_size)
return
# Histogram plot
bins = [0.0]
binWidth = self.options.cd_bin_width
binEnd = binWidth
while binEnd <= 1.0:
bins.append(binEnd)
binEnd += binWidth
axesHist.hist(data, bins=bins, normed=True, color=(0.5, 0.5, 0.5))
axesHist.set_xlabel('% coding density')
axesHist.set_ylabel('% windows (' + str(self.options.cd_window_size) + ' bp)')
# Prettify plot
for a in axesHist.yaxis.majorTicks:
a.tick1On = True
a.tick2On = False
for a in axesHist.xaxis.majorTicks:
a.tick1On = True
a.tick2On = False
for line in axesHist.yaxis.get_ticklines():
line.set_color(self.axesColour)
for line in axesHist.xaxis.get_ticklines():
line.set_color(self.axesColour)
for loc, spine in axesHist.spines.iteritems():
if loc in ['right', 'top']:
spine.set_color('none')
else:
spine.set_color(self.axesColour)
# get CD bin statistics
binTools = BinTools()
meanCD, deltaCDs, _ = binTools.codingDensityDist(seqs, prodigalParser)
# Delta-CD vs sequence length plot
axesDeltaCD.scatter(deltaCDs, seqLens, c=abs(deltaCDs), s=10, lw=0.5, cmap='gray_r')
axesDeltaCD.set_xlabel(r'$\Delta$ CD (mean coding density = %.1f%%)' % (meanCD * 100))
axesDeltaCD.set_ylabel('Sequence length (kbp)')
_, yMaxSeqs = axesDeltaCD.get_ylim()
xMinSeqs, xMaxSeqs = axesDeltaCD.get_xlim()
# plot reference distributions
for distToPlot in distributionsToPlot:
closestCD = findNearest(np.array(dist.keys()), meanCD)
# find closest distribution values
sampleSeqLen = dist[closestCD].keys()[0]
d = dist[closestCD][sampleSeqLen]
cdLowerBoundKey = findNearest(d.keys(), (100 - distToPlot) / 2.0)
cdUpperBoundKey = findNearest(d.keys(), (100 + distToPlot) / 2.0)
xL = []
xU = []
y = []
for windowSize in dist[closestCD]:
xL.append(dist[closestCD][windowSize][cdLowerBoundKey])
xU.append(dist[closestCD][windowSize][cdUpperBoundKey])
y.append(windowSize)
# sort by y-values
sortIndexY = np.argsort(y)
xL = np.array(xL)[sortIndexY]
xU = np.array(xU)[sortIndexY]
y = np.array(y)[sortIndexY]
axesDeltaCD.plot(xL, y, 'r--', lw=0.5, zorder=0)
axesDeltaCD.plot(xU, y, 'r--', lw=0.5, zorder=0)
# ensure y-axis include zero and covers all sequences
axesDeltaCD.set_ylim([0, yMaxSeqs])
# ensure x-axis is set appropriately for sequences
axesDeltaCD.set_xlim([xMinSeqs, xMaxSeqs])
# draw vertical line at x=0
axesDeltaCD.vlines(0, 0, yMaxSeqs, linestyle='dashed', color=self.axesColour, zorder=0)
# Change sequence lengths from bp to kbp
yticks = axesDeltaCD.get_yticks()
kbpLabels = []
for seqLen in yticks:
label = '%.1f' % (float(seqLen) / 1000)
label = label.replace('.0', '') # remove trailing zero
kbpLabels.append(label)
axesDeltaCD.set_yticklabels(kbpLabels)
# Prettify plot
for a in axesDeltaCD.yaxis.majorTicks:
a.tick1On = True
a.tick2On = False
for a in axesDeltaCD.xaxis.majorTicks:
a.tick1On = True
a.tick2On = False
for line in axesDeltaCD.yaxis.get_ticklines():
line.set_color(self.axesColour)
for line in axesDeltaCD.xaxis.get_ticklines():
line.set_color(self.axesColour)
for loc, spine in axesDeltaCD.spines.iteritems():
if loc in ['right', 'top']:
spine.set_color('none')
else:
spine.set_color(self.axesColour)
def identifyOutliers(self, outDir, binFiles, tetraProfileFile,
distribution, reportType, outputFile):
"""Identify sequences that are outliers."""
self.logger.info('Reading reference distributions.')
gcBounds = readDistribution('gc_dist')
cdBounds = readDistribution('cd_dist')
tdBounds = readDistribution('td_dist')
fout = open(outputFile, 'w')
fout.write(
'Bin Id\tSequence Id\tSequence length\tOutlying distributions')
fout.write(
'\tSequence GC\tMean bin GC\tLower GC bound (%s%%)\tUpper GC bound (%s%%)'
% (distribution, distribution))
fout.write('\tSequence CD\tMean bin CD\tLower CD bound (%s%%)' %
distribution)
fout.write('\tSequence TD\tMean bin TD\tUpper TD bound (%s%%)\n' %
distribution)
processedBins = 0
for binFile in binFiles:
binId = binIdFromFilename(binFile)
processedBins += 1
self.logger.info('Finding outliers in %s (%d of %d).' %
(binId, processedBins, len(binFiles)))
seqs = readFasta(binFile)
meanGC, deltaGCs, seqGC = self.gcDist(seqs)
genomicSig = GenomicSignatures(K=4, threads=1)
tetraSigs = genomicSig.read(tetraProfileFile)
binSig = self.binTetraSig(seqs, tetraSigs)
meanTD, deltaTDs = self.tetraDiffDist(seqs, genomicSig, tetraSigs,
binSig)
gffFile = os.path.join(outDir, 'bins', binId,
DefaultValues.PRODIGAL_GFF)
if not os.path.exists(gffFile):
self.logger.error(
'Missing gene feature file (%s). This plot if not compatible with the --genes option.\n'
% DefaultValues.PRODIGAL_GFF)
sys.exit(1)
prodigalParser = ProdigalGeneFeatureParser(gffFile)
meanCD, deltaCDs, CDs = self.codingDensityDist(
seqs, prodigalParser)
# find keys into GC and CD distributions
closestGC = findNearest(np.array(list(gcBounds.keys())), meanGC)
sampleSeqLen = list(gcBounds[closestGC].keys())[0]
d = gcBounds[closestGC][sampleSeqLen]
gcLowerBoundKey = findNearest(list(d.keys()),
(100 - distribution) / 2.0)
gcUpperBoundKey = findNearest(list(d.keys()),
(100 + distribution) / 2.0)
closestCD = findNearest(np.array(list(cdBounds.keys())), meanCD)
sampleSeqLen = list(cdBounds[closestCD].keys())[0]
d = cdBounds[closestCD][sampleSeqLen]
cdLowerBoundKey = findNearest(list(d.keys()),
(100 - distribution) / 2.0)
tdBoundKey = findNearest(
list(tdBounds[list(tdBounds.keys())[0]].keys()), distribution)
index = 0
for seqId, seq in seqs.items():
seqLen = len(seq)
# find GC, CD, and TD bounds
closestSeqLen = findNearest(list(gcBounds[closestGC].keys()),
seqLen)
gcLowerBound = gcBounds[closestGC][closestSeqLen][
gcLowerBoundKey]
gcUpperBound = gcBounds[closestGC][closestSeqLen][
gcUpperBoundKey]
closestSeqLen = findNearest(list(cdBounds[closestCD].keys()),
seqLen)
cdLowerBound = cdBounds[closestCD][closestSeqLen][
cdLowerBoundKey]
closestSeqLen = findNearest(list(tdBounds.keys()), seqLen)
tdBound = tdBounds[closestSeqLen][tdBoundKey]
outlyingDists = []
if deltaGCs[index] < gcLowerBound or deltaGCs[
index] > gcUpperBound:
outlyingDists.append('GC')
if deltaCDs[index] < cdLowerBound:
outlyingDists.append('CD')
if deltaTDs[index] > tdBound:
outlyingDists.append('TD')
if (reportType == 'any' and len(outlyingDists) >= 1) or (
reportType == 'all' and len(outlyingDists) == 3):
fout.write(binId + '\t' + seqId + '\t%d' % len(seq) +
'\t' + ','.join(outlyingDists))
fout.write('\t%.1f\t%.1f\t%.1f\t%.1f' %
(seqGC[index] * 100, meanGC * 100,
(meanGC + gcLowerBound) * 100,
(meanGC + gcUpperBound) * 100))
fout.write('\t%.1f\t%.1f\t%.1f' %
(CDs[index] * 100, meanCD * 100,
(meanCD + cdLowerBound) * 100))
fout.write('\t%.3f\t%.3f\t%.3f' %
(deltaTDs[index], meanTD, tdBound) + '\n')
index += 1
fout.close()
def __calculateResults(self, windowSizes, numWindows, genomeDir, queueIn, queueOut):
while True:
genomeId = queueIn.get(block=True, timeout=None)
if genomeId == None:
break
seqs = readGenomicSeqsFromFasta(os.path.join(genomeDir, genomeId, genomeId + '.fna'))
# for simplicity, create a single scaffold from all sequences
genomeFile = os.path.join('./deltaCD/genomes', genomeId + '.single_scaffold.fna')
genomeScaffold = 'NNNNNNNNNN'.join(seqs.values()).upper()
fout = open(genomeFile, 'w')
fout.write('>' + genomeId + '\n')
fout.write(genomeScaffold)
fout.close()
# run prodigal on genome
ntFile = os.path.join('./deltaCD/prodigal', genomeId + '.genes.fna')
gffFile = os.path.join('./deltaCD/prodigal', genomeId + '.gff')
cmd = ('prodigal -q -c -m -f gff -d %s -i %s > %s' % (ntFile, genomeFile, gffFile))
os.system(cmd)
# calculate mean coding density of genome
numericScaffold = self.__createNumericScaffold(genomeScaffold)
prodigalParser = ProdigalGeneFeatureParser(gffFile)
codingBases = prodigalParser.codingBases(genomeId)
counts = np.bincount(numericScaffold)
totalBases = counts[0]
meanCD = float(codingBases) / totalBases
fout = open('./deltaCD/' + genomeId + '.tsv', 'w')
fout.write('# Mean CD = ' + str(meanCD) + '\n')
# calculate coding density distribution for different window sizes
for windowSize in windowSizes:
endWindowPos = len(genomeScaffold) - windowSize
if endWindowPos <= 0:
# This might occur for the largest window sizes and smallest genomes
break
deltaCDs = []
while len(deltaCDs) != numWindows:
# pick random window
startWindow = randint(0, endWindowPos)
# calculate coding density
codingBases = prodigalParser.codingBases(genomeId, startWindow, startWindow+windowSize)
counts = np.bincount(numericScaffold[startWindow:(startWindow+windowSize)])
totalBases = counts[0]
if totalBases != windowSize:
# there are N's in the window so skip it
continue
cdPer = float(codingBases) / totalBases
deltaCDs.append(cdPer - meanCD)
fout.write('Windows Size = ' + str(windowSize) + '\n')
fout.write(','.join(map(str, deltaCDs)) + '\n')
fout.close()
queueOut.put(genomeId)
