def test_sampleLabelings(self):
tree = Tree("(A:1,(B:1,(C:1,(E:1,D:1)Int_1:0.5[&&NHX:ancient=1])Int_2:0.5[&&NHX:ancient=0])Int_3:1)Root;", format=1)
chrom = {}
chrom["one"] = ["3","4"]
species = {}
species["C"] = chrom
chrom = {}
chrom["one"] = ["3","4"]
species["D"] = chrom
chrom = {}
chrom["one"] = []
species["E"] = chrom
chrom = {}
chrom["one"] = []
species["A"] = chrom
chrom = {}
chrom["one"] = []
species["B"] = chrom
adj = getAdjacencies.findAdjacencies(species)
paths = getAdjacencies.findTreePaths(tree)
internal,adjacenciesAncestral = getAdjacencies.assignAncestralAdjacencies(paths,adj,tree)
graphs = globalAdjacencyGraph.createGraph(adj,adjacenciesAncestral)
jointLabels, first = SR.enumJointLabelings(graphs)
probs={"Int_1":{(6, 7):0.1},"Int_2":{(6, 7):0.1},"Int_3":{(6, 7):0.1},"Root":{(6, 7):0.1}}
for i in range(0,10):
validLabels, validAtNode = SR.validLabels(jointLabels,first)
resolvedCCs = SR.sampleLabelings(tree, graphs, validAtNode, adj,probs, alpha=0)
reconstructedAdj = SR.reconstructedAdjacencies(resolvedCCs)
print reconstructedAdj
def test_enumJointLabelings(self):
#print nx.maximal_matching(self.graph)
#print nx.max_weight_matching(self.graph)
joint,first = SR.enumJointLabelings([self.graph])
TestCase.assertEqual(self,len(joint[self.graph]),15)
valid, validAtNode = SR.validLabels(joint,first)
TestCase.assertEqual(self,len(valid[self.graph]),8)
def runSample(params):
#retrieving the given parameter
ccs=params[0]
tree=params[1]
extantAdjacencies=params[2]
adjacencyProbs=params[3]
alpha=params[4]
i=params[5]
extantAdjacencies_species_adj=params[6]
outputDirectory=params[7]
reconstructedMarkerCount=params[8]
allSampleReconstructionStatistic={}
dict_SCJ={}
#lock = multiprocessing.Lock()
#output text log
outLog="Sample: "+str(i)+"\n"
#start sampling method like in the Main.py
outLog+="Enumerate joint labelings...\n"
jointLabels, first = SR.enumJointLabelings(ccs)
outLog+="Check valid labels...\n"
validLabels, validAtNode = SR.validLabels(jointLabels, first)
#lock.acquire()
outLog+= "Compute ancestral labels with SR...\n"
topDown = SR.sampleLabelings(tree, ccs, validAtNode, extantAdjacencies, adjacencyProbs, alpha)
#lock.release()
reconstructedAdj = SR.reconstructedAdjacencies(topDown)
SR.outputReconstructedAdjacencies(reconstructedAdj, outputDirectory+"/reconstructed_adjacencies_" + str(i))
for node in reconstructedAdj:
# count for each adjaency on each internal node, how often this adjacencies over all samples occurs there
for adjacency in reconstructedAdj[node]:
#lock.acquire()
if (node,adjacency) in allSampleReconstructionStatistic:
allSampleReconstructionStatistic[(node,adjacency)] += 1
else:
allSampleReconstructionStatistic.update({(node,adjacency):1})
#lock.release()
outLog+="Scaffolding...\n"
scaffolds = scaffolding.scaffoldAdjacencies(reconstructedAdj)
undoubled = scaffolding.undoubleScaffolds(scaffolds)
scaffolding.outputUndoubledScaffolds(undoubled, outputDirectory+"/undoubled_scaffolds_" + str(i))
scaffolding.outputScaffolds(scaffolds, outputDirectory+"/doubled_scaffolds_" + str(i))
log=scaffolding.sanityCheckScaffolding(undoubled)
outLog+=log
for node in undoubled:
outLog+= str(node)+'\n'
markerCounter = 0
for scaffold in undoubled[node]:
first = scaffold[0]
last = scaffold[-1]
if not first == last:
markerCounter = markerCounter + len(scaffold)
else:
markerCounter = markerCounter + len(scaffold) - 1
outLog+= str(node) + " number of reconstructed undoubled marker in scaffolds: " + str(markerCounter)+'\n'
# number of reconstructed markerIds given by reconstructedMarkerCount
# singleton scaffolds number / number of not reconstructed marker
notReconstructedMarkerCount = reconstructedMarkerCount - markerCounter
# number of all scaffolds
allScaffoldCount = len(undoubled[node]) + notReconstructedMarkerCount
outLog+= str(node) + " number of singleton scaffolds (not reconstructed marker): " + str(
notReconstructedMarkerCount)+'\n'
outLog+= str(node) + " number of scaffolds: " + str(allScaffoldCount)+'\n'
#lock.acquire()
scj = calculate_SCJ(tree, reconstructedAdj, extantAdjacencies_species_adj)
outLog+="Single-Cut-or-Join-Distance: " + str(scj)+'\n'
dict_SCJ.update({'Sample_' + str(i): scj})
#lock.release()
return (allSampleReconstructionStatistic,dict_SCJ,outLog)
adjacencyProbs[species]={adj:weight}
line=f.readline()
f.close()
#dictionary for all scj distances
dict_SCJ={}
#compute CCs in global adjacency graph
ccs = globalAdjacencyGraph.createGraph(extantAdjacencies,nodesPerAdjacency)
if (not args.skip_first):
conflicts = globalAdjacencyGraph.analyseConnectedComponents(ccs)
globalAdjacencyGraph.outputConflicts(conflicts,args.output+"/conflicts")
jointLabels, first = SR.enumJointLabelings(ccs)
validLabels, validAtNode = SR.validLabels(jointLabels,first)
topDown = SR.computeLabelings(tree, ccs, validAtNode, extantAdjacencies, adjacencyProbs, args.alpha)
reconstructedAdj = SR.reconstructedAdjacencies(topDown)
SR.outputReconstructedAdjacencies(reconstructedAdj,args.output+"/reconstructed_adjacencies")
for node in reconstructedAdj:
print node
print "Number of reconstructed adjacencies: "+str(len(reconstructedAdj[node]))
scaffolds = scaffolding.scaffoldAdjacencies(reconstructedAdj)
undoubled = scaffolding.undoubleScaffolds(scaffolds)
scaffolding.outputUndoubledScaffolds(undoubled,args.output+"/undoubled_scaffolds")
scaffolding.outputScaffolds(scaffolds,args.output+"/doubled_scaffolds")
scaffolding.sanityCheckScaffolding(undoubled)