def formSeqGraph(folderName, mummerLink):
print "formSeqGraph"
startList, graphNodes = [], []
print "formSeqGraph: Reading in best successors and predecessors"
rightConnect = readConnectList(folderName, "rightConnect.txt")
leftConnect = readConnectList(folderName, "leftConnect.txt")
numberOfNodes = len(rightConnect)
print "formSeqGraph: Initializing seqGraph"
G = graphLib.seqGraph(numberOfNodes)
print "formSeqGraph: Adding edges to seqGraph"
for eachitem, i in zip(rightConnect, range(len(rightConnect))):
index = i
connector, weight = eachitem
G.insertEdge(index, connector, weight)
for eachitem, i in zip(leftConnect, range(len(leftConnect))):
index = i
connector, weight = eachitem
G.insertEdge(connector, index, weight)
G.cleanEdge()
G.condense()
print "formSeqGraph: Outputting seqGraph to condensedGraph.txt"
G.saveToFile(folderName, "condensedGraph.txt")
G.checkSelfLoops()
G.checkCompleteness()
G2 = graphLib.seqGraph(0)
G2.loadFromFile(folderName, "condensedGraph.txt")
houseKeeper.compareGraphUnitTest(G, G2)
G.reportDummyUsefulNode()
G.reportEdge()
graphFileName = "condensedGraph.txt"
contigFile = "noEmbed_Double.fasta"
outContigFile = "improved.fasta"
outOpenList = "openZone.txt"
print "formSeqGraph: Outputting improved contigs from seqGraph to improved.fasta"
IORobot.readContigOut(folderName, mummerLink, graphFileName, contigFile,
outContigFile, outOpenList)
def readContigOut(folderName, mummerLink, graphFileName, contigFile,
outContigFile, outOpenList):
print "readContigOut"
G = graphLib.seqGraph(0)
G.loadFromFile(folderName, graphFileName)
G.findStartEndList()
myContigsDic = loadContigsFromFile(folderName, contigFile)
contigUsed = [False for i in range(len(G.graphNodesList) / 2)]
seqToPrint = []
openList = []
noForRevMismatch = True
for eachnode in G.graphNodesList:
if len(eachnode.nodeIndexList) > 0:
tmpSeq = ""
# ## debug consistency of t/f
ckList = []
for dummy in eachnode.nodeIndexList:
indexToAdd = dummy
readNum = indexToAdd / 2
ckList.append(contigUsed[readNum])
if (len(ckList) > 0 and not all(ckList) and any(ckList)):
noForRevMismatch = False
# ## end debug
for i in range(len(eachnode.nodeIndexList)):
indexToAdd = eachnode.nodeIndexList[i]
readNum = indexToAdd / 2
orientation = indexToAdd % 2
if contigUsed[readNum] == False:
if i != len(eachnode.nodeIndexList) - 1:
overlapLen = eachnode.overlapList[i]
if orientation == 0:
tmpSeq = tmpSeq + myContigsDic['Contig' +
str(readNum) + '_' +
'p'][0:-overlapLen]
else:
tmpSeq = tmpSeq + myContigsDic['Contig' +
str(readNum) + '_' +
'd'][0:-overlapLen]
else:
if orientation == 0:
tmpSeq = tmpSeq + myContigsDic['Contig' +
str(readNum) + '_' +
'p']
else:
tmpSeq = tmpSeq + myContigsDic['Contig' +
str(readNum) + '_' +
'd']
contigUsed[readNum] = True
if len(tmpSeq) > 0:
if eachnode.nodeIndex in G.myStartList:
openList.append('Segkk' + str(len(seqToPrint)) + ',noprev')
if eachnode.nodeIndex in G.myEndList:
openList.append('Segkk' + str(len(seqToPrint)) + ',nonext')
# ## Debug
if eachnode.nodeIndex == 444:
print 439, len(seqToPrint)
if eachnode.nodeIndex == 67:
print 67, len(seqToPrint)
# ## End Debug
seqToPrint.append(tmpSeq)
print "No forward/reverse mismatch ?", noForRevMismatch
fImproved = open(folderName + outContigFile, 'w')
for eachcontig, dummyIndex in zip(seqToPrint, range(len(seqToPrint))):
fImproved.write(">Segkk" + str(dummyIndex) + '\n')
fImproved.write(eachcontig + '\n')
fImproved.close()
print "All contigs used? ", all(contigUsed)
print "NContig", len(seqToPrint)
f = open(folderName + outOpenList, 'w')
f.write(str(len(seqToPrint)) + '\n')
for eachitem in openList:
f.write(str(eachitem) + str('\n'))
f.close()
def xPhased(folderName , mummerLink):
# ## Repeat resolution [Proxy for MB]
# 1. Re-form the contig string graph with ALL connections from contigs only V
# 2. Log down the reads and associated blocked contigs V
# 3. Use reads to connect;
# 4. Transform graph by identifying 1 successor/predecessor case ; Condense(important);
# 5. Read out contigs
print "xPhased: Aligning improved2.fasta against itself, outputting to mb*.delta"
numberOfContig, dataSet = IORobot.obtainLinkInfo(folderName, mummerLink, "improved2", "mb")
lenDic = IORobot.obtainLength(folderName, "improved2_Double.fasta")
confidenLenThres = 0
print "xPhased: Building seqGraph"
G = graphLib.seqGraph(numberOfContig)
extraEdges = loadEdgeFromBlockedReads(folderName)
for eachitem in dataSet:
# print eachitem
wt, myin, myout = eachitem
myInData = myin[6:].split('_')
myOutData = myout[6:].split('_')
if myInData[1] == 'p':
offsetin = 0
else:
offsetin = 1
if myOutData[1] == 'p':
offsetout = 0
else:
offsetout = 1
i = int(myInData[0]) * 2 + offsetin
j = int(myOutData[0]) * 2 + offsetout
ck = False
for eachedge in extraEdges:
mystart, myend, len1, len2 = eachedge[0], eachedge[1], eachedge[2] , eachedge[3]
if [i, j] == [mystart, myend] and min(len1, len2) >= wt and lenDic[myin] >= confidenLenThres and lenDic[myout] >= confidenLenThres:
ck = True
if ck:
G.insertEdge(i, j, wt)
# G.reportEdge()
G.MBResolve()
G.reportEdge()
print "xPhased: Saving condensed seqGraph to condensedGraphMB.txt"
G.saveToFile(folderName, "condensedGraphMB.txt")
graphFileName = "condensedGraphMB.txt"
contigFile = "improved2_Double.fasta"
outContigFile = "improved3.fasta"
outOpenList = "openZoneMB.txt"
print "xPhased: Outputting improved contigs from condensed seqGraph to improved3.fasta"
IORobot.readContigOut(folderName, mummerLink, graphFileName, contigFile, outContigFile, outOpenList)
# ## Repeat resolution [Proxy for phasing step]
# 6. Find out the repeat region by MSA
# 7. Find out the location of SNPs and extend across repeat
# [short cut : use contig creator : your job here is to get data into the correct formats]
print "xPhased"
def readContigOut(folderName,
mummerLink,
graphFileName,
contigFile,
outContigFile,
outOpenList,
nameDic={}):
print "readContigOut"
G = graphLib.seqGraph(0)
G.loadFromFile(folderName, graphFileName)
G.findStartEndList()
myContigsDic = loadContigsFromFile(folderName, contigFile)
contigUsed = [False for i in range(len(G.graphNodesList) / 2)]
seqToPrint = []
openList = []
noForRevMismatch = True
print len(G.graphNodesList)
for eachnode in G.graphNodesList:
print eachnode.nodeIndexList
if len(eachnode.nodeIndexList) > 0:
tmpSeq = ""
# ## debug consistency of t/f
ckList = []
for dummy in eachnode.nodeIndexList:
indexToAdd = dummy
readNum = indexToAdd / 2
ckList.append(contigUsed[readNum])
if (len(ckList) > 0 and not all(ckList) and any(ckList)):
noForRevMismatch = False
# ## end debug
if contigUsed[eachnode.nodeIndexList[0] / 2] == False:
contigUsed[eachnode.nodeIndexList[0] / 2] = True
contigUsed[eachnode.nodeIndexList[-1] / 2] = True
for i in range(len(eachnode.nodeIndexList)):
indexToAdd = eachnode.nodeIndexList[i]
readNum = indexToAdd / 2
orientation = indexToAdd % 2
#print nameDic[indexToAdd]
if len(nameDic) > 0:
orientation = nameDic[indexToAdd] % 2
readNum = nameDic[indexToAdd] / 2
#print readNum
if i != len(eachnode.nodeIndexList) - 1:
overlapLenOld = eachnode.overlapList[i]
# Can we hijack here for the overlap Length... seems like minimal changes
overlapLen = useAlignToGetLen(eachnode, i, nameDic,
orientation,
myContigsDic, readNum,
folderName, mummerLink)
# End Hijacking
print overlapLen, overlapLenOld
if orientation == 0:
tmpSeq = tmpSeq + myContigsDic['Contig' +
str(readNum) + '_' +
'p'][0:-overlapLen]
else:
tmpSeq = tmpSeq + myContigsDic['Contig' +
str(readNum) + '_' +
'd'][0:-overlapLen]
else:
if orientation == 0:
tmpSeq = tmpSeq + myContigsDic['Contig' +
str(readNum) + '_' +
'p']
else:
tmpSeq = tmpSeq + myContigsDic['Contig' +
str(readNum) + '_' +
'd']
if len(tmpSeq) > 0:
if eachnode.nodeIndex in G.myStartList:
openList.append('Segkk' + str(len(seqToPrint)) + ',noprev')
if eachnode.nodeIndex in G.myEndList:
openList.append('Segkk' + str(len(seqToPrint)) + ',nonext')
seqToPrint.append(tmpSeq)
print "No forward/reverse mismatch ?", noForRevMismatch
fImproved = open(folderName + outContigFile, 'w')
for eachcontig, dummyIndex in zip(seqToPrint, range(len(seqToPrint))):
print len(eachcontig)
fImproved.write(">Segkk" + str(dummyIndex) + '\n')
fImproved.write(eachcontig + '\n')
fImproved.close()
print "All contigs used? ", all(contigUsed)
print "NContig", len(seqToPrint)
f = open(folderName + outOpenList, 'w')
f.write(str(len(seqToPrint)) + '\n')
for eachitem in openList:
f.write(str(eachitem) + str('\n'))
f.close()
