def run(self, minThreshold, maxThreshold, stepSize, minGenomes, mostSpecificRanks):
img = IMG()
trustedGenomeIds = img.trustedGenomes()
fout = open("./data/markerSetSize.tsv", "w")
fout.write("Lineage\t# genomes")
for threshold in arange(maxThreshold, minThreshold, -stepSize):
fout.write("\t" + str(threshold))
fout.write("\n")
lineages = img.lineagesSorted(mostSpecificRanks)
for lineage in lineages:
genomeIds = img.genomeIdsByTaxonomy(lineage)
genomeIds = list(genomeIds.intersection(trustedGenomeIds))
if len(genomeIds) < minGenomes:
continue
print "\nLineage " + lineage + " contains " + str(len(genomeIds)) + " genomes."
fout.write(lineage + "\t" + str(len(genomeIds)))
pfamTable = img.pfamTable(genomeIds)
for threshold in arange(maxThreshold, minThreshold, -stepSize):
markerSet = img.markerGenes(
genomeIds, pfamTable, threshold * len(genomeIds), threshold * len(genomeIds)
)
fout.write("\t" + str(len(markerSet)))
print " Threshold = %.2f, marker set size = %d" % (threshold, len(markerSet))
fout.write("\n")
fout.close()
def run(self, taxonomyStr, minThreshold, maxThreshold, stepSize):
img = IMG()
genomeIds = img.genomeIdsByTaxonomy(taxonomyStr, 'Final')
print('Lineage ' + taxonomyStr + ' contains ' + str(len(genomeIds)) + ' genomes.')
markerSetSizes = []
countTable = img.countTable(genomeIds)
for threshold in arange(maxThreshold, minThreshold, -stepSize):
markerGenes = img.markerGenes(genomeIds, countTable, threshold*len(genomeIds), threshold*len(genomeIds))
geneDistTable = img.geneDistTable(genomeIds, markerGenes, spacingBetweenContigs=1e6)
colocatedGenes = img.colocatedGenes(geneDistTable)
colocatedSets = img.colocatedSets(colocatedGenes, markerGenes)
markerSetSizes.append(len(colocatedSets))
print(' Threshold = %.2f, marker set size = %d' % (threshold, len(markerGenes)))
# plot data
plot = LinePlot()
plotFilename = './images/markerSetSize.' + taxonomyStr.replace(';','_') + '.png'
title = taxonomyStr.replace(';', '; ')
plot.plot(plotFilename, arange(maxThreshold, minThreshold, -stepSize), markerSetSizes, 'Threshold', 'Marker Set Size', title)
def run(self, minThreshold, maxThreshold, stepSize, minGenomes,
mostSpecificRanks):
img = IMG()
trustedGenomeIds = img.trustedGenomes()
fout = open('./data/markerSetSize.tsv', 'w')
fout.write('Lineage\t# genomes')
for threshold in arange(maxThreshold, minThreshold, -stepSize):
fout.write('\t' + str(threshold))
fout.write('\n')
lineages = img.lineagesSorted(mostSpecificRanks)
for lineage in lineages:
genomeIds = img.genomeIdsByTaxonomy(lineage)
genomeIds = list(genomeIds.intersection(trustedGenomeIds))
if len(genomeIds) < minGenomes:
continue
print('\nLineage ' + lineage + ' contains ' + str(len(genomeIds)) +
' genomes.')
fout.write(lineage + '\t' + str(len(genomeIds)))
pfamTable = img.pfamTable(genomeIds)
for threshold in arange(maxThreshold, minThreshold, -stepSize):
markerSet = img.markerGenes(genomeIds, pfamTable,
threshold * len(genomeIds),
threshold * len(genomeIds))
fout.write('\t' + str(len(markerSet)))
print(' Threshold = %.2f, marker set size = %d' %
(threshold, len(markerSet)))
fout.write('\n')
fout.close()
def run(self, taxonomyStr, minThreshold, maxThreshold, stepSize):
img = IMG()
genomeIds = img.genomeIdsByTaxonomy(taxonomyStr, 'Final')
print 'Lineage ' + taxonomyStr + ' contains ' + str(len(genomeIds)) + ' genomes.'
markerSetSizes = []
countTable = img.countTable(genomeIds)
for threshold in arange(maxThreshold, minThreshold, -stepSize):
markerGenes = img.markerGenes(genomeIds, countTable, threshold*len(genomeIds), threshold*len(genomeIds))
geneDistTable = img.geneDistTable(genomeIds, markerGenes, spacingBetweenContigs=1e6)
colocatedGenes = img.colocatedGenes(geneDistTable)
colocatedSets = img.colocatedSets(colocatedGenes, markerGenes)
markerSetSizes.append(len(colocatedSets))
print ' Threshold = %.2f, marker set size = %d' % (threshold, len(markerGenes))
# plot data
plot = LinePlot()
plotFilename = './images/markerSetSize.' + taxonomyStr.replace(';','_') + '.png'
title = taxonomyStr.replace(';', '; ')
plot.plot(plotFilename, arange(maxThreshold, minThreshold, -stepSize), markerSetSizes, 'Threshold', 'Marker Set Size', title)
def run(self, ubiquityThreshold, singleCopyThreshold, minGenomes, minMarkers, mostSpecificRank, percentGenomes, numReplicates):
img = IMG()
lineages = img.lineagesByCriteria(minGenomes, mostSpecificRank)
fout = open('./data/lineage_evaluation.tsv', 'w')
fout.write('Lineage\t# genomes\t# markers\tpercentage\tnum replicates\tmean\tstd\tmean %\tmean + std%\tmean + 2*std %\n')
for lineage in lineages:
genomeIds = img.genomeIdsByTaxonomy(lineage, 'Final')
if len(genomeIds) < minGenomes:
continue
countTable = img.countTable(genomeIds)
countTable = img.filterTable(genomeIds, countTable, ubiquityThreshold*0.9, singleCopyThreshold*0.9)
# calculate marker set for all genomes
markerGenes = img.markerGenes(genomeIds, countTable, ubiquityThreshold*len(genomeIds), singleCopyThreshold*len(genomeIds))
if len(markerGenes) < minMarkers:
continue
print '\nLineage ' + lineage + ' contains ' + str(len(genomeIds)) + ' genomes.'
print ' Marker genes: ' + str(len(markerGenes))
fout.write(lineage + '\t' + str(len(genomeIds)) + '\t' + str(len(markerGenes)) + '\t%.2f' % percentGenomes + '\t' + str(numReplicates))
# withhold select percentage of genomes and calculate new marker set
changeMarkerSetSize = []
for _ in xrange(0, numReplicates):
subsetGenomeIds = random.sample(genomeIds, int((1.0-percentGenomes)*len(genomeIds) + 0.5))
newMarkerGenes = img.markerGenes(subsetGenomeIds, countTable, ubiquityThreshold*len(subsetGenomeIds), singleCopyThreshold*len(subsetGenomeIds))
changeMarkerSetSize.append(len(newMarkerGenes.symmetric_difference(markerGenes)))
m = mean(changeMarkerSetSize)
s = std(changeMarkerSetSize)
print ' Mean: %.2f, Std: %.2f, Per: %.2f' % (m, s, (m+ 2*s) * 100 / len(markerGenes))
fout.write('\t%.2f\t%.2f\t%.2f\t%.2f\t%.2f' % (m, s, m * 100 / len(markerGenes), (m + s) * 100 / len(markerGenes), (m + 2*s) * 100 / len(markerGenes)) + '\n')
fout.close()
def run(self, taxonomyStr, ubiquityThreshold, singleCopyThreshold,
percentCompletion, numReplicates, numGenomes, contigLen):
img = IMG()
genomeIds = img.genomeIdsByTaxonomy(taxonomyStr, 'Final')
print('\nLineage ' + taxonomyStr + ' contains ' + str(len(genomeIds)) +
' genomes.')
# build marker genes and colocated marker sets
countTable = img.countTable(genomeIds)
markerGenes = img.markerGenes(genomeIds, countTable,
ubiquityThreshold * len(genomeIds),
singleCopyThreshold * len(genomeIds))
print(' Marker genes: ' + str(len(markerGenes)))
geneDistTable = img.geneDistTable(genomeIds,
markerGenes,
spacingBetweenContigs=1e6)
colocatedGenes = img.colocatedGenes(geneDistTable)
colocatedSets = img.colocatedSets(colocatedGenes, markerGenes)
print(' Co-located gene sets: ' + str(len(colocatedSets)))
# random sample genomes
if numGenomes == -1:
rndGenomeIds = genomeIds
else:
rndGenomeIds = random.sample(genomeIds, numGenomes)
# estimate completion for each genome using both the marker genes and marker sets
metadata = img.genomeMetadata('Final')
plotLabels = []
plotData = []
for genomeId in rndGenomeIds:
mgCompletion = []
msCompletion = []
for _ in range(0, numReplicates):
startPartialGenomeContigs = img.sampleGenome(
metadata[genomeId]['genome size'], percentCompletion,
contigLen)
# calculate completion with marker genes
containedMarkerGenes = img.containedMarkerGenes(
markerGenes, geneDistTable[genomeId],
startPartialGenomeContigs, contigLen)
mgCompletion.append(
float(len(containedMarkerGenes)) / len(markerGenes) -
percentCompletion)
# calculate completion with marker set
comp = 0.0
for cs in colocatedSets:
present = 0
for contigId in cs:
if contigId in containedMarkerGenes:
present += 1
comp += float(present) / len(cs)
msCompletion.append(comp / len(colocatedSets) -
percentCompletion)
plotData.append(mgCompletion)
plotData.append(msCompletion)
species = ' '.join(
metadata[genomeId]['taxonomy'][ranksByLabel['Genus']:])
plotLabels.append(species + ' (' + genomeId + ')')
plotLabels.append('')
# plot data
boxPlot = BoxPlot()
plotFilename = './images/sim.MGvsMS.' + taxonomyStr.replace(
';', '_') + '.' + str(percentCompletion) + '.errorbar.png'
title = taxonomyStr.replace(
';', '; ') + '\n' + 'Percent completion = %.2f' % percentCompletion
boxPlot.plot(plotFilename, plotData, plotLabels,
r'$\Delta$' + ' Percent Completion', '', False, title)
def run(self, taxonomyStr, mostSpecificRank, minGenomes, ubiquityThreshold,
singleCopyThreshold, percentCompletion, numReplicates, numGenomes,
contigLen):
img = IMG()
lineages = []
taxon = taxonomyStr.split(';')
for r in range(0, len(taxon)):
lineages.append(';'.join(taxon[0:r + 1]))
# get all marker sets
markerGenes = []
geneDistTable = []
colocatedSets = []
for lineage in lineages:
genomeIds = img.genomeIdsByTaxonomy(lineage, 'Final')
print('\nLineage ' + lineage + ' contains ' + str(len(genomeIds)) +
' genomes.')
# build marker genes and colocated marker sets
countTable = img.countTable(genomeIds)
mg = img.markerGenes(genomeIds, countTable,
ubiquityThreshold * len(genomeIds),
singleCopyThreshold * len(genomeIds))
print(' Marker genes: ' + str(len(mg)))
mdt = img.geneDistTable(genomeIds, mg, spacingBetweenContigs=1e6)
colocatedGenes = img.colocatedGenes(mdt)
cs = img.colocatedSets(colocatedGenes, mg)
print(' Co-located gene sets: ' + str(len(cs)))
markerGenes.append(mg)
geneDistTable.append(mdt)
colocatedSets.append(cs)
# random sample genomes
if numGenomes == -1:
rndGenomeIds = genomeIds
else:
rndGenomeIds = random.sample(genomeIds, numGenomes)
# estimate completion for each genome using both the marker genes and marker sets
metadata = img.genomeMetadata('Final')
plotLabels = []
plotData = []
for genomeId in rndGenomeIds:
completion = [[] for _ in range(len(lineages))]
for _ in range(0, numReplicates):
startPartialGenomeContigs = img.sampleGenome(
metadata[genomeId]['genome size'], percentCompletion,
contigLen)
# calculate completion with marker set
for i in range(len(lineages)):
containedMarkerGenes = img.containedMarkerGenes(
markerGenes[i], geneDistTable[i][genomeId],
startPartialGenomeContigs, contigLen)
comp = 0.0
for cs in colocatedSets[i]:
present = 0
for contigId in cs:
if contigId in containedMarkerGenes:
present += 1
comp += float(present) / len(cs)
completion[i].append(comp / len(colocatedSets[i]) -
percentCompletion)
plotLabels.append(genomeId + ' - ' + lineages[i])
for d in completion:
plotData.append(d)
# plot data
boxPlot = BoxPlot()
plotFilename = './images/sim.lineages.' + taxonomyStr.replace(
';', '_') + '.' + str(percentCompletion) + '.errorbar.png'
title = taxonomyStr.replace(
';', '; ') + '\n' + 'Percent completion = %.2f' % percentCompletion
boxPlot.plot(plotFilename, plotData, plotLabels,
r'$\Delta$' + ' Percent Completion', '', False, title)
def run(self, ubiquityThreshold, singleCopyThreshold, minGenomes,
minMarkers, mostSpecificRank, percentGenomes, numReplicates):
img = IMG()
lineages = img.lineagesByCriteria(minGenomes, mostSpecificRank)
fout = open('./data/lineage_evaluation.tsv', 'w')
fout.write(
'Lineage\t# genomes\t# markers\tpercentage\tnum replicates\tmean\tstd\tmean %\tmean + std%\tmean + 2*std %\n'
)
for lineage in lineages:
genomeIds = img.genomeIdsByTaxonomy(lineage, 'Final')
if len(genomeIds) < minGenomes:
continue
countTable = img.countTable(genomeIds)
countTable = img.filterTable(genomeIds, countTable,
ubiquityThreshold * 0.9,
singleCopyThreshold * 0.9)
# calculate marker set for all genomes
markerGenes = img.markerGenes(genomeIds, countTable,
ubiquityThreshold * len(genomeIds),
singleCopyThreshold * len(genomeIds))
if len(markerGenes) < minMarkers:
continue
print '\nLineage ' + lineage + ' contains ' + str(
len(genomeIds)) + ' genomes.'
print ' Marker genes: ' + str(len(markerGenes))
fout.write(lineage + '\t' + str(len(genomeIds)) + '\t' +
str(len(markerGenes)) + '\t%.2f' % percentGenomes +
'\t' + str(numReplicates))
# withhold select percentage of genomes and calculate new marker set
changeMarkerSetSize = []
for _ in xrange(0, numReplicates):
subsetGenomeIds = random.sample(
genomeIds,
int((1.0 - percentGenomes) * len(genomeIds) + 0.5))
newMarkerGenes = img.markerGenes(
subsetGenomeIds, countTable,
ubiquityThreshold * len(subsetGenomeIds),
singleCopyThreshold * len(subsetGenomeIds))
changeMarkerSetSize.append(
len(newMarkerGenes.symmetric_difference(markerGenes)))
m = mean(changeMarkerSetSize)
s = std(changeMarkerSetSize)
print ' Mean: %.2f, Std: %.2f, Per: %.2f' % (m, s,
(m + 2 * s) * 100 /
len(markerGenes))
fout.write('\t%.2f\t%.2f\t%.2f\t%.2f\t%.2f' %
(m, s, m * 100 / len(markerGenes),
(m + s) * 100 / len(markerGenes),
(m + 2 * s) * 100 / len(markerGenes)) + '\n')
fout.close()
def run(self, ubiquityThreshold, singleCopyThreshold, minGenomes,
mostSpecificRank, minMarkers):
print('Ubiquity threshold: ' + str(ubiquityThreshold))
print('Single-copy threshold: ' + str(singleCopyThreshold))
print('Min. genomes: ' + str(minGenomes))
print('Most specific taxonomic rank: ' + str(mostSpecificRank))
img = IMG()
deltaMarkerSetSizes = []
lineages = img.lineagesByCriteria(minGenomes, mostSpecificRank)
lineages = ['prokaryotes'] + lineages
boxPlotLabels = []
for lineage in lineages:
genomeIds = img.genomeIdsByTaxonomy(lineage)
trusted = img.trustedGenomes()
genomeIds = list(genomeIds.intersection(trusted))
print('')
print('Lineage ' + lineage + ' contains ' + str(len(genomeIds)) +
' genomes.')
# get table of PFAMs and do some initial filtering to remove PFAMs that are
# clearly not going to pass the ubiquity and single-copy thresholds
pfamTable = img.pfamTable(genomeIds)
pfamTable = img.filterPfamTable(genomeIds, pfamTable,
ubiquityThreshold * 0.9,
singleCopyThreshold * 0.9)
markerSet = img.markerGenes(
genomeIds, pfamTable, ubiquityThreshold * (len(genomeIds) - 1),
singleCopyThreshold * (len(genomeIds) - 1))
fullMarkerSetSize = len(markerSet)
if fullMarkerSetSize < minMarkers:
continue
boxPlotLabels.append(
lineage.split(';')[-1].strip() + ' (' + str(len(genomeIds)) +
', ' + str(fullMarkerSetSize) + ')')
deltaMarkerSetSize = []
numGenomes = len(genomeIds) - 1
for loo in range(0, len(genomeIds)):
if loo != len(genomeIds) - 1:
genomeIdSubset = genomeIds[0:loo] + genomeIds[loo + 1:]
else:
genomeIdSubset = genomeIds[0:loo]
markerSet = img.markerGenes(
genomeIdSubset, pfamTable,
ubiquityThreshold * len(genomeIdSubset),
singleCopyThreshold * len(genomeIdSubset))
deltaMarkerSetSize.append(fullMarkerSetSize - len(markerSet))
if fullMarkerSetSize < len(markerSet):
print('[Warning] Unexpected!')
deltaMarkerSetSizes.append(deltaMarkerSetSize)
m = mean(deltaMarkerSetSize)
s = std(deltaMarkerSetSize)
print(' LOO Ubiquity >= ' +
str(int(ubiquityThreshold * numGenomes)) +
', LOO Single-copy >= ' +
str(int(singleCopyThreshold * numGenomes)))
print(' Delta Mean: %.2f +/- %.2f' % (m, s))
print(' Delta Min: %d, Delta Max: %d' %
(min(deltaMarkerSetSize), max(deltaMarkerSetSize)))
# plot data
boxPlot = BoxPlot()
plotFilename = './images/LOO.' + str(ubiquityThreshold) + '-' + str(
singleCopyThreshold) + '.boxplot.png'
title = 'Ubiquity = %.2f' % ubiquityThreshold + ', Single-copy = %.2f' % singleCopyThreshold
boxPlot.plot(plotFilename, deltaMarkerSetSizes, boxPlotLabels,
r'$\Delta$' + ' Marker Set Size', '', False, title)
def run(self, taxonomyStr, ubiquityThreshold, singleCopyThreshold, percentCompletion, numReplicates, numGenomes, contigLen):
img = IMG()
genomeIds = img.genomeIdsByTaxonomy(taxonomyStr, 'Final')
print '\nLineage ' + taxonomyStr + ' contains ' + str(len(genomeIds)) + ' genomes.'
# build marker genes and colocated marker sets
countTable = img.countTable(genomeIds)
markerGenes = img.markerGenes(genomeIds, countTable, ubiquityThreshold*len(genomeIds), singleCopyThreshold*len(genomeIds))
print ' Marker genes: ' + str(len(markerGenes))
geneDistTable = img.geneDistTable(genomeIds, markerGenes)
colocatedGenes = img.colocatedGenes(geneDistTable)
colocatedSets = img.colocatedSets(colocatedGenes, markerGenes)
print ' Co-located gene sets: ' + str(len(colocatedSets))
# random sample genomes
if numGenomes == -1:
rndGenomeIds = genomeIds
else:
rndGenomeIds = random.sample(genomeIds, numGenomes)
# estimate completion for each genome using both the marker genes and marker sets
metadata = img.genomeMetadata('Final')
plotLabels = []
plotData = []
for genomeId in rndGenomeIds:
mgCompletion = []
msCompletion = []
for _ in xrange(0, numReplicates):
startPartialGenomeContigs = img.sampleGenome(metadata[genomeId]['genome size'], percentCompletion, contigLen)
# calculate completion with marker genes
containedMarkerGenes = img.containedMarkerGenes(markerGenes, geneDistTable[genomeId], startPartialGenomeContigs, contigLen)
mgCompletion.append(float(len(containedMarkerGenes))/len(markerGenes) - percentCompletion)
# calculate completion with marker set
comp = 0.0
for cs in colocatedSets:
present = 0
for contigId in cs:
if contigId in containedMarkerGenes:
present += 1
comp += float(present) / len(cs)
msCompletion.append(comp / len(colocatedSets) - percentCompletion)
plotData.append(mgCompletion)
plotData.append(msCompletion)
species = ' '.join(metadata[genomeId]['taxonomy'][ranksByLabel['Genus']:])
plotLabels.append(species + ' (' + genomeId + ')')
plotLabels.append('')
# plot data
boxPlot = BoxPlot()
plotFilename = './images/sim.MGvsMS.' + taxonomyStr.replace(';','_') + '.' + str(percentCompletion) + '.errorbar.png'
title = taxonomyStr.replace(';', '; ') + '\n' + 'Percent completion = %.2f' % percentCompletion
boxPlot.plot(plotFilename, plotData, plotLabels, r'$\Delta$' + ' Percent Completion', '', False, title)
def run(self, ubiquityThreshold, singleCopyThreshold, minGenomes, mostSpecificRank, minMarkers):
print 'Ubiquity threshold: ' + str(ubiquityThreshold)
print 'Single-copy threshold: ' + str(singleCopyThreshold)
print 'Min. genomes: ' + str(minGenomes)
print 'Most specific taxonomic rank: ' + str(mostSpecificRank)
img = IMG()
deltaMarkerSetSizes = []
lineages = img.lineagesByCriteria(minGenomes, mostSpecificRank)
lineages = ['prokaryotes'] + lineages
boxPlotLabels = []
for lineage in lineages:
genomeIds = img.genomeIdsByTaxonomy(lineage)
trusted = img.trustedGenomes()
genomeIds = list(genomeIds.intersection(trusted))
print ''
print 'Lineage ' + lineage + ' contains ' + str(len(genomeIds)) + ' genomes.'
# get table of PFAMs and do some initial filtering to remove PFAMs that are
# clearly not going to pass the ubiquity and single-copy thresholds
pfamTable = img.pfamTable(genomeIds)
pfamTable = img.filterPfamTable(genomeIds, pfamTable, ubiquityThreshold*0.9, singleCopyThreshold*0.9)
markerSet = img.markerGenes(genomeIds, pfamTable, ubiquityThreshold*(len(genomeIds)-1), singleCopyThreshold*(len(genomeIds)-1))
fullMarkerSetSize = len(markerSet)
if fullMarkerSetSize < minMarkers:
continue
boxPlotLabels.append(lineage.split(';')[-1].strip() + ' (' + str(len(genomeIds)) + ', ' + str(fullMarkerSetSize) + ')')
deltaMarkerSetSize = []
numGenomes = len(genomeIds)-1
for loo in xrange(0, len(genomeIds)):
if loo != len(genomeIds) - 1:
genomeIdSubset = genomeIds[0:loo] + genomeIds[loo+1:]
else:
genomeIdSubset = genomeIds[0:loo]
markerSet = img.markerGenes(genomeIdSubset, pfamTable, ubiquityThreshold*len(genomeIdSubset), singleCopyThreshold*len(genomeIdSubset))
deltaMarkerSetSize.append(fullMarkerSetSize - len(markerSet))
if fullMarkerSetSize < len(markerSet):
print '[Warning] Unexpected!'
deltaMarkerSetSizes.append(deltaMarkerSetSize)
m = mean(deltaMarkerSetSize)
s = std(deltaMarkerSetSize)
print ' LOO Ubiquity >= ' + str(int(ubiquityThreshold*numGenomes)) + ', LOO Single-copy >= ' + str(int(singleCopyThreshold*numGenomes))
print ' Delta Mean: %.2f +/- %.2f' % (m, s)
print ' Delta Min: %d, Delta Max: %d' % (min(deltaMarkerSetSize), max(deltaMarkerSetSize))
# plot data
boxPlot = BoxPlot()
plotFilename = './images/LOO.' + str(ubiquityThreshold) + '-' + str(singleCopyThreshold) + '.boxplot.png'
title = 'Ubiquity = %.2f' % ubiquityThreshold + ', Single-copy = %.2f' % singleCopyThreshold
boxPlot.plot(plotFilename, deltaMarkerSetSizes, boxPlotLabels, r'$\Delta$' + ' Marker Set Size', '', False, title)
def run(self, taxonomyStr, mostSpecificRank, minGenomes, ubiquityThreshold, singleCopyThreshold, percentCompletion, numReplicates, numGenomes, contigLen):
img = IMG()
lineages = []
taxon = taxonomyStr.split(';')
for r in xrange(0, len(taxon)):
lineages.append(';'.join(taxon[0:r+1]))
# get all marker sets
markerGenes = []
geneDistTable = []
colocatedSets = []
for lineage in lineages:
genomeIds = img.genomeIdsByTaxonomy(lineage, 'Final')
print '\nLineage ' + lineage + ' contains ' + str(len(genomeIds)) + ' genomes.'
# build marker genes and colocated marker sets
countTable = img.countTable(genomeIds)
mg = img.markerGenes(genomeIds, countTable, ubiquityThreshold*len(genomeIds), singleCopyThreshold*len(genomeIds))
print ' Marker genes: ' + str(len(mg))
mdt = img.geneDistTable(genomeIds, mg, spacingBetweenContigs=1e6)
colocatedGenes = img.colocatedGenes(mdt)
cs = img.colocatedSets(colocatedGenes, mg)
print ' Co-located gene sets: ' + str(len(cs))
markerGenes.append(mg)
geneDistTable.append(mdt)
colocatedSets.append(cs)
# random sample genomes
if numGenomes == -1:
rndGenomeIds = genomeIds
else:
rndGenomeIds = random.sample(genomeIds, numGenomes)
# estimate completion for each genome using both the marker genes and marker sets
metadata = img.genomeMetadata('Final')
plotLabels = []
plotData = []
for genomeId in rndGenomeIds:
completion = [[] for _ in xrange(len(lineages))]
for _ in xrange(0, numReplicates):
startPartialGenomeContigs = img.sampleGenome(metadata[genomeId]['genome size'], percentCompletion, contigLen)
# calculate completion with marker set
for i in xrange(len(lineages)):
containedMarkerGenes = img.containedMarkerGenes(markerGenes[i], geneDistTable[i][genomeId], startPartialGenomeContigs, contigLen)
comp = 0.0
for cs in colocatedSets[i]:
present = 0
for contigId in cs:
if contigId in containedMarkerGenes:
present += 1
comp += float(present) / len(cs)
completion[i].append(comp / len(colocatedSets[i]) - percentCompletion)
plotLabels.append(genomeId + ' - ' + lineages[i])
for d in completion:
plotData.append(d)
# plot data
boxPlot = BoxPlot()
plotFilename = './images/sim.lineages.' + taxonomyStr.replace(';','_') + '.' + str(percentCompletion) + '.errorbar.png'
title = taxonomyStr.replace(';', '; ') + '\n' + 'Percent completion = %.2f' % percentCompletion
boxPlot.plot(plotFilename, plotData, plotLabels, r'$\Delta$' + ' Percent Completion', '', False, title)
