def logEdges(self, folderName, stagename):
print "Logging edges"
logList = []
if stagename == "XResolution":
mapDummyToRealDic =self.readInJSON(folderName, "mapDummyToRealDic.json")
else:
mapDummyToRealDic = {}
for eachnode in self.graphNodesList:
tmpNodeIndexList = []
for kk in eachnode.nodeIndexList:
if kk >= self.N1:
tmpNodeIndexList += mapDummyToRealDic[str(kk-self.N1)][1]
else:
tmpNodeIndexList += [kk]
if len(tmpNodeIndexList) >= 2:
for i in range(len(tmpNodeIndexList)-1):
currentName = tmpNodeIndexList[i]
nextName = tmpNodeIndexList[i+1]
cName = abunHouseKeeper.parseIDToName(currentName,'C',0)
nName = abunHouseKeeper.parseIDToName(nextName,'C',0)
logList.append([cName, nName])
with open( folderName + stagename + ".json", 'w') as f:
json.dump(logList, f)
def getCtAgg(inList, myCountDic, Gnew, lenDic):
newInList = []
print "New getCtAgg", len(inList)
for i in inList:
tmp1 = convert4to1base(i)
tmp2 = convert4to2base(i)
covTmp = 0
lenTmp = 0
print "len(Gnew.graphNodesList[tmp2].nodeIndexList), len(inList)", len(
Gnew.graphNodesList[tmp2].nodeIndexList), len(inList)
for eachindex in Gnew.graphNodesList[tmp2].nodeIndexList:
if lenDic[abunHouseKeeper.parseIDToName(eachindex, 'C',
len(lenDic))] > lenTmp:
lenTmp = lenDic[abunHouseKeeper.parseIDToName(
eachindex, 'C', len(lenDic))]
name = "Segkk" + str(eachindex / 2)
covTmp = myCountDic[name]
newInList.append([tmp2, covTmp])
return newInList
def condenseEdgeRemove(self, G_ContigRead, folderName, mummerLink, contigFilename):
print "condenseEdgeRemove"
thresPass = 100
thresForStrangeCut = 5000
### kkdebug
toRemoveList = []
for eachnode in self.graphNodesList:
if len(eachnode.nodeIndexList) > 0:
if len(eachnode.listOfNextNodes) ==1 :
nextNodeIndex = eachnode.listOfNextNodes[0][0]
nextNode= self.graphNodesList[nextNodeIndex]
if len(nextNode.listOfPrevNodes) == 1 :
currentName = eachnode.nodeIndex
nextName = nextNode.nodeIndex
contigReadPaths = findAllPathK(currentName,nextName, G_ContigRead, 5)
cName = abunHouseKeeper.parseIDToName(currentName,'C',0)
nName = abunHouseKeeper.parseIDToName(nextName,'C',0)
noGoNext = self.readInJSON(folderName, "noGoNext.json")
noGoPrev = self.readInJSON(folderName, "noGoPrev.json")
overlap = [-1, -1]
ctr = 0
for eachpath in contigReadPaths:
if len(eachpath) > 2:
ctr = ctr + 1
elif len(eachpath) == 2:
contigName = cName
leftSeg = IORobot.myRead(folderName, contigFilename + "_Double.fasta", contigName)
contigName = nName
rightSeg = IORobot.myRead(folderName, contigFilename + "_Double.fasta", contigName)
overlap = IORobot.align(leftSeg, rightSeg, folderName, mummerLink)
if ctr <= thresPass and (cName in noGoNext or nName in noGoPrev or overlap[0] > thresForStrangeCut ):
self.removeEdge(currentName, nextName)
toRemoveList.append([currentName, nextName])
### kkdebug
#with open( "dataFolder/toRemoveList.json", 'w') as f:
# json.dump(toRemoveList, f)
self.findAdjList()
def condenseEdgeRemove(self, G_ContigRead, folderName, mummerLink, contigFilename):
print "condenseEdgeRemove"
thresPass = 100
thresForStrangeCut = 5000
### kkdebug
toRemoveList = []
for eachnode in self.graphNodesList:
if len(eachnode.nodeIndexList) > 0:
if len(eachnode.listOfNextNodes) ==1 :
nextNodeIndex = eachnode.listOfNextNodes[0][0]
nextNode= self.graphNodesList[nextNodeIndex]
if len(nextNode.listOfPrevNodes) == 1 :
currentName = eachnode.nodeIndex
nextName = nextNode.nodeIndex
contigReadPaths = findAllPathK(currentName,nextName, G_ContigRead, 5)
cName = abunHouseKeeper.parseIDToName(currentName,'C',0)
nName = abunHouseKeeper.parseIDToName(nextName,'C',0)
noGoNext = self.readInJSON(folderName, "noGoNext.json")
noGoPrev = self.readInJSON(folderName, "noGoPrev.json")
overlap = [-1, -1]
ctr = 0
for eachpath in contigReadPaths:
if len(eachpath) > 2:
ctr = ctr + 1
elif len(eachpath) == 2:
contigName = cName
leftSeg = IORobot.myRead(folderName, contigFilename + "_Double.fasta", contigName)
contigName = nName
rightSeg = IORobot.myRead(folderName, contigFilename + "_Double.fasta", contigName)
overlap = IORobot.align(leftSeg, rightSeg, folderName, mummerLink)
if ctr <= thresPass and (cName in noGoNext or nName in noGoPrev or overlap[0] > thresForStrangeCut ):
self.removeEdge(currentName, nextName)
toRemoveList.append([currentName, nextName])
### kkdebug
#with open( "dataFolder/toRemoveList.json", 'w') as f:
# json.dump(toRemoveList, f)
self.findAdjList()
def findNoGoByNoHeads(noGoList, side, folderName):
noGoListNew = []
sortedContigList, sortedReadList, sortedContigDic, sortedReadDic =\
formSortedDataList(folderName)
lenDicContig = IORobot.obtainLength(folderName, "mFixed_Double.fasta")
lenDicRead = IORobot.obtainLength(folderName,
"phasingSeedName_Double.fasta")
for x in noGoList:
rList = findAttachedReads(x, side, folderName, sortedContigList,
sortedContigDic, lenDicContig, lenDicRead)
cList = findAttachedContigs(rList, side, folderName, sortedReadList,
sortedReadDic, lenDicContig, lenDicRead)
if bestMatchContigOnly == False:
bestContigIDList = findBreakContigAdv(cList)
else:
bestContigIDList = findBreakContig(cList)
if len(rList) > 0 and len(cList) > 0:
print "x, side, len(rList), len(cList), len(bestContigIDList)",\
abunHouseKeeper.parseIDToName(x,'C',0), side, len(rList), len(cList), len(bestContigIDList)
print "cList", bestContigIDList
noGoListNew = noGoListNew + bestContigIDList
return noGoListNew
def findAttachedReads(x, side, folderName, sortedContigList, sortedContigDic,
lenDicContig, lenDicRead):
rList = []
'''
Format :
[S1] [E1] | [S2] [E2] | [LEN 1] [LEN 2] | [ IDY] | [TAGS]
=====================================================================================
1 562 | 819 1418 | 562 600 | 84.72 | Contig0_d Read121_d
1 562 | 4077 3478 | 562 600 | 84.72 | Contig0_d Read121_p
1 564 | 656 68 | 564 589 | 90.13 | Contig0_d Read382_d
1 564 | 6996 7584 | 564 589 | 90.13 | Contig0_d Read382_p
1 571 | 1386 815 | 571 572 | 86.60 | Contig0_d Read421_d
'''
thres = thresMiddleContig
key = abunHouseKeeper.parseIDToName(x, 'C', 0)
if key in sortedContigDic:
tmp = sortedContigDic[key]
while tmp < len(sortedContigList) and sortedContigList[tmp][-2] == key:
eachsub = sortedContigList[tmp]
if overlapCR(eachsub, side, thres, lenDicContig, lenDicRead):
rList.append(eachsub[-1])
tmp = tmp + 1
distinctRList = abunHouseKeeper.getDistinct(rList)
else:
distinctRList = []
return distinctRList
def findNoGoByNoHeads(noGoList, side, folderName): noGoListNew = [] sortedContigList, sortedReadList, sortedContigDic, sortedReadDic =\ formSortedDataList(folderName) lenDicContig = IORobot.obtainLength(folderName, "mFixed_Double.fasta" ) lenDicRead = IORobot.obtainLength(folderName, "phasingSeedName_Double.fasta") for x in noGoList: rList = findAttachedReads(x, side, folderName,sortedContigList,sortedContigDic, lenDicContig,lenDicRead) cList = findAttachedContigs(rList, side, folderName, sortedReadList, sortedReadDic, lenDicContig,lenDicRead) if bestMatchContigOnly == False: bestContigIDList = findBreakContigAdv(cList) else: bestContigIDList = findBreakContig(cList) if len(rList) > 0 and len(cList) > 0: print "x, side, len(rList), len(cList), len(bestContigIDList)",\ abunHouseKeeper.parseIDToName(x,'C',0), side, len(rList), len(cList), len(bestContigIDList) print "cList", bestContigIDList noGoListNew = noGoListNew + bestContigIDList return noGoListNew
def bipartiteLocalResolve(self, resolvedList, inList, outList, folderName):
#noGoNext = self.readInJSON(folderName, "noGoNext.json")
#noGoPrev = self.readInJSON(folderName, "noGoPrev.json")
for u in inList:
self.clearOut(u/2)
for v in outList:
self.clearIn(v/2)
for e in resolvedList:
u, v =e[0], e[-1]
cName = abunHouseKeeper.parseIDToName(u,'C',0)
nName = abunHouseKeeper.parseIDToName(v,'C',0)
#if not cName in noGoNext and not nName in noGoPrev:
self.insertEdge(u,v,1997)
def bipartiteLocalResolve(self, resolvedList, inList, outList, folderName):
#noGoNext = self.readInJSON(folderName, "noGoNext.json")
#noGoPrev = self.readInJSON(folderName, "noGoPrev.json")
if len(resolvedList) > 0:
for u in inList:
self.clearOut(u/2)
for v in outList:
self.clearIn(v/2)
for e in resolvedList:
u, v =e[0], e[-1]
cName = abunHouseKeeper.parseIDToName(u,'C',0)
nName = abunHouseKeeper.parseIDToName(v,'C',0)
#if not cName in noGoNext and not nName in noGoPrev:
self.insertEdge(u,v,1997)
def getCtAgg(inList, myCountDic, Gnew, lenDic):
newInList = []
print "New getCtAgg", len(inList)
for i in inList:
tmp1 = convert4to1base(i)
tmp2 = convert4to2base(i)
covTmp = 0
lenTmp = 0
print "len(Gnew.graphNodesList[tmp2].nodeIndexList), len(inList)", len(
Gnew.graphNodesList[tmp2].nodeIndexList
), len(inList)
for eachindex in Gnew.graphNodesList[tmp2].nodeIndexList:
if lenDic[abunHouseKeeper.parseIDToName(eachindex, "C", len(lenDic))] > lenTmp:
lenTmp = lenDic[abunHouseKeeper.parseIDToName(eachindex, "C", len(lenDic))]
name = "Segkk" + str(eachindex / 2)
covTmp = myCountDic[name]
newInList.append([tmp2, covTmp])
return newInList
def findAttachedReads(x, side, folderName,sortedContigList,sortedContigDic, lenDicContig,lenDicRead):
rList = []
'''
Format :
[S1] [E1] | [S2] [E2] | [LEN 1] [LEN 2] | [ IDY] | [TAGS]
=====================================================================================
1 562 | 819 1418 | 562 600 | 84.72 | Contig0_d Read121_d
1 562 | 4077 3478 | 562 600 | 84.72 | Contig0_d Read121_p
1 564 | 656 68 | 564 589 | 90.13 | Contig0_d Read382_d
1 564 | 6996 7584 | 564 589 | 90.13 | Contig0_d Read382_p
1 571 | 1386 815 | 571 572 | 86.60 | Contig0_d Read421_d
'''
thres = thresMiddleContig
key = abunHouseKeeper.parseIDToName(x, 'C', 0)
if key in sortedContigDic:
tmp = sortedContigDic[key]
while tmp < len(sortedContigList) and sortedContigList[tmp][-2] == key:
eachsub = sortedContigList[tmp]
if overlapCR(eachsub, side, thres, lenDicContig,lenDicRead):
rList.append(eachsub[-1])
tmp = tmp + 1
distinctRList = abunHouseKeeper.getDistinct(rList)
else:
distinctRList = []
return distinctRList
def singleGapLookUp(eachmatchpair,folderName, N1, mummerLink, contigReadGraph, contigFilename,readsetFilename):
#print eachmatchpair
leftCtgIndex ,rightCtgIndex, leftEnd, rightStart, middleContent = eachmatchpair[0],eachmatchpair[-1],0,0,""
succReadsList = []
G = seqGraphWt(0)
G.loadFromFile(folderName, contigReadGraph)
succReadsList = BFS(leftCtgIndex,rightCtgIndex, G, N1)
if len(succReadsList) > 0:
succReadsList.pop(0)
succReadsList.pop(-1)
else:
print "interesting item for future study"
print "succReadsList" , succReadsList
if len(succReadsList) == 0:
contigName = abunHouseKeeper.parseIDToName(leftCtgIndex, 'C', N1)
leftSeg = IORobot.myRead(folderName, contigFilename + "_Double.fasta", contigName)
contigName = abunHouseKeeper.parseIDToName(rightCtgIndex, 'C', N1)
rightSeg = IORobot.myRead(folderName, contigFilename + "_Double.fasta", contigName)
overlap = IORobot.alignWithName(leftSeg, rightSeg, folderName, mummerLink, str(leftCtgIndex) + "_" + str(rightCtgIndex) )
print "overlap contig : ", overlap
leftEnd = len(leftSeg) - overlap[0]
middleContent = ""
else:
contigName = abunHouseKeeper.parseIDToName(leftCtgIndex, 'C', N1)
print contigName
leftSeg = IORobot.myRead(folderName, contigFilename + "_Double.fasta", contigName)
readName = abunHouseKeeper.parseIDToName(succReadsList[0], 'R', N1)
print readName
rightSeg = IORobot.myRead(folderName, readsetFilename + "_Double.fasta", readName)
overlap = IORobot.alignWithName(leftSeg, rightSeg, folderName, mummerLink, str(leftCtgIndex) + "_" + str(rightCtgIndex) )
print "overlap start read : ", overlap
leftEnd = len(leftSeg) - overlap[0]
middleContent = ""
for i in range(len(succReadsList)-1):
readName = abunHouseKeeper.parseIDToName(succReadsList[i], 'R', N1)
leftSeg = IORobot.myRead(folderName, readsetFilename + "_Double.fasta", readName)
readName = abunHouseKeeper.parseIDToName(succReadsList[i+1], 'R', N1)
rightSeg = IORobot.myRead(folderName, readsetFilename + "_Double.fasta", readName)
overlap = IORobot.alignWithName(leftSeg, rightSeg, folderName, mummerLink, str(leftCtgIndex) + "_" + str(rightCtgIndex) )
print "overlap middle read : ", overlap
middleContent = middleContent + leftSeg[0:len(leftSeg)-overlap[0]]
readName = abunHouseKeeper.parseIDToName(succReadsList[-1], 'R', N1)
leftSeg = IORobot.myRead(folderName, readsetFilename + "_Double.fasta", readName)
contigName = abunHouseKeeper.parseIDToName(rightCtgIndex, 'C', N1)
rightSeg = IORobot.myRead(folderName, contigFilename + "_Double.fasta", contigName)
overlap = IORobot.alignWithName(leftSeg, rightSeg, folderName, mummerLink, str(leftCtgIndex) + "_" + str(rightCtgIndex) )
print "overlap end read : ", overlap
middleContent = middleContent + leftSeg[0:len(leftSeg)-overlap[0]]
return [leftCtgIndex ,rightCtgIndex, leftEnd, rightStart, middleContent]
def singleGapLookUp(eachmatchpair, folderName, N1, mummerLink, contigReadGraph, contigFilename, readsetFilename):
print eachmatchpair
leftCtgIndex, rightCtgIndex, leftEnd, rightStart, middleContent = eachmatchpair[0], eachmatchpair[-1], 0, 0, ""
succReadsList = abunGraphLib.findPathBtwEnds(folderName, leftCtgIndex, rightCtgIndex, contigReadGraph, N1)
succReadsList = []
G = graphLib.seqGraph(0)
G.loadFromFile(folderName, contigReadGraph)
allPaths = abunGraphLib.findAllPathK(leftCtgIndex, rightCtgIndex, G, 5)
# shuffle(allPaths)
print "allPaths", allPaths
possibleList = []
for p in allPaths:
noContig = True
for pp in p[1:-1]:
if pp < N1:
noContig = False
if noContig == True:
possibleList.append(p)
print "possibleList", possibleList
minListLen = 1000
for p in possibleList:
if len(p) < minListLen:
succReadsList = p
minListLen = len(p)
if len(succReadsList) > 0:
succReadsList.pop(0)
succReadsList.pop(-1)
else:
print "interesting item for future study"
print "succReadsList", succReadsList
if len(succReadsList) == 0:
contigName = abunHouseKeeper.parseIDToName(leftCtgIndex, "C", N1)
leftSeg = IORobot.myRead(folderName, contigFilename + "_Double.fasta", contigName)
contigName = abunHouseKeeper.parseIDToName(rightCtgIndex, "C", N1)
rightSeg = IORobot.myRead(folderName, contigFilename + "_Double.fasta", contigName)
overlap = IORobot.alignWithName(
leftSeg, rightSeg, folderName, mummerLink, str(leftCtgIndex) + "_" + str(rightCtgIndex)
)
print "overlap contig : ", overlap
leftEnd = len(leftSeg) - overlap[0]
middleContent = ""
else:
contigName = abunHouseKeeper.parseIDToName(leftCtgIndex, "C", N1)
print contigName
leftSeg = IORobot.myRead(folderName, contigFilename + "_Double.fasta", contigName)
readName = abunHouseKeeper.parseIDToName(succReadsList[0], "R", N1)
print readName
rightSeg = IORobot.myRead(folderName, readsetFilename + "_Double.fasta", readName)
overlap = IORobot.alignWithName(
leftSeg, rightSeg, folderName, mummerLink, str(leftCtgIndex) + "_" + str(rightCtgIndex)
)
print "overlap start read : ", overlap
leftEnd = len(leftSeg) - overlap[0]
middleContent = ""
for i in range(len(succReadsList) - 1):
readName = abunHouseKeeper.parseIDToName(succReadsList[i], "R", N1)
leftSeg = IORobot.myRead(folderName, readsetFilename + "_Double.fasta", readName)
readName = abunHouseKeeper.parseIDToName(succReadsList[i + 1], "R", N1)
rightSeg = IORobot.myRead(folderName, readsetFilename + "_Double.fasta", readName)
overlap = IORobot.alignWithName(
leftSeg, rightSeg, folderName, mummerLink, str(leftCtgIndex) + "_" + str(rightCtgIndex)
)
print "overlap middle read : ", overlap
middleContent = middleContent + leftSeg[0 : len(leftSeg) - overlap[0]]
readName = abunHouseKeeper.parseIDToName(succReadsList[-1], "R", N1)
leftSeg = IORobot.myRead(folderName, readsetFilename + "_Double.fasta", readName)
contigName = abunHouseKeeper.parseIDToName(rightCtgIndex, "C", N1)
rightSeg = IORobot.myRead(folderName, contigFilename + "_Double.fasta", contigName)
overlap = IORobot.alignWithName(
leftSeg, rightSeg, folderName, mummerLink, str(leftCtgIndex) + "_" + str(rightCtgIndex)
)
print "overlap end read : ", overlap
middleContent = middleContent + leftSeg[0 : len(leftSeg) - overlap[0]]
return [leftCtgIndex, rightCtgIndex, leftEnd, rightStart, middleContent]