def adjListToRepeatList(newAdjacencyList, folderName, repeatFilename):
N1 = len(newAdjacencyList)
G2 = abunGraphLib.seqGraphWt(N1 * 2)
for i in range(N1):
for j in newAdjacencyList[i]:
G2.insertEdge(2 * i, 2 * j + 1, 1)
G2.insertEdge(2 * j + 1, 2 * i, 1)
clusters = G2.findConnectedComponents()
repeatList = []
for eachitem in clusters:
leftList, rightList = [], []
for eachsubitem in eachitem:
if eachsubitem % 2 == 0:
leftList.append(eachsubitem)
else:
rightList.append(eachsubitem)
repeatList.append([
abunHouseKeeper.getDistinct(leftList),
abunHouseKeeper.getDistinct(rightList)
])
with open(folderName + repeatFilename, 'w') as outfile:
json.dump(repeatList, outfile)
json_data = open(folderName + repeatFilename, 'r')
loadData = json.load(json_data)
assert (loadData == repeatList)
def formConfirmReadResolve(folderName, inList, outList, G, Grev, N1):
#print "formConfirmReadResolve"
resolvedList = []
confirmingReadList = []
brLFlankList = []
brRFlankList = []
### Find possible candidate reads
print "inList , outList formConfirmReadResolve()", inList, outList
for eachin in inList:
for eachout in outList:
pathList = abunGraphLib.findAllPathK(eachin, eachout, G, 3)
for path in pathList:
if len(path) == 3 and path[1] >= N1:
R = path[1]
confirmingReadList.append(R)
brLFlankList.append([eachin, R])
brRFlankList.append([eachout, R])
### Filter simple false cases
toUseReadDic = {}
confirmingReadList.sort()
for key, items in groupby(confirmingReadList):
toUseReadDic[str(key)] = True
newbrLFlankList = abunHouseKeeper.getDistinct(brLFlankList)
newbrLFlankList.sort(key=itemgetter(1))
for key, items in groupby(newbrLFlankList, itemgetter(1)):
mylist = list(items)
if len(mylist) > 1:
toUseReadDic[str(key)] = False
newbrRFlankList = abunHouseKeeper.getDistinct(brRFlankList)
newbrRFlankList.sort(key=itemgetter(1))
for key, items in groupby(newbrRFlankList, itemgetter(1)):
mylist = list(items)
if len(mylist) > 1:
toUseReadDic[str(key)] = False
finalSearchReadList = []
for eachitem in toUseReadDic:
if toUseReadDic[eachitem] == True:
finalSearchReadList.append(int(eachitem))
### Check paths to confirm all false cases
for eachR in finalSearchReadList:
l1 = abunGraphLib.findAllReachable(eachR, N1, G)
l2 = abunGraphLib.findAllReachable(eachR, N1, Grev)
l1Distinct = abunHouseKeeper.getDistinct(l1)
l2Distinct = abunHouseKeeper.getDistinct(l2)
if len(l1Distinct) == 1 and len(l2Distinct) == 1:
c1, c2 = l1Distinct[0], l2Distinct[0]
resolvedList.append([c2, c1])
return resolvedList
def adjListToRepeatList(newAdjacencyList,folderName,repeatFilename):
N1 = len(newAdjacencyList)
G2 = abunGraphLib.seqGraphWt(N1 * 2)
for i in range(N1):
for j in newAdjacencyList[i]:
G2.insertEdge(2 * i, 2 * j + 1, 1)
G2.insertEdge(2 * j + 1, 2 * i, 1)
clusters = G2.findConnectedComponents()
repeatList = []
for eachitem in clusters:
leftList, rightList = [], []
for eachsubitem in eachitem:
if eachsubitem % 2 == 0 :
leftList.append(eachsubitem)
else:
rightList.append(eachsubitem)
repeatList.append([abunHouseKeeper.getDistinct(leftList), abunHouseKeeper.getDistinct(rightList)])
with open(folderName + repeatFilename, 'w') as outfile:
json.dump(repeatList, outfile)
json_data = open(folderName + repeatFilename, 'r')
loadData = json.load(json_data)
assert(loadData == repeatList)
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 filterConfidResolve(resolvedList):
newResolvedList = []
resolvedList.sort()
conThres = abunHouseKeeper.abunGlobalSplitParameterRobot.BRThres
print "conThres", conThres
for key, items in groupby(resolvedList):
tmpList = list(items)
if len(tmpList) >= conThres:
newResolvedList.append(key)
if False:
noConflict = resolveConflict(abunHouseKeeper.getDistinct(resolvedList))
noConflict = abunHouseKeeper.getDistinct(noConflict)
newResolvedList = abunHouseKeeper.getDistinct(newResolvedList)
newResolvedList = abunHouseKeeper.getDistinct(intersect(newResolvedList, noConflict))
return newResolvedList
def resolveConflictX(listA, listB):
resolvedList = [[] for i in range(len(listA))]
print "len(listA), len(listB)", len(listA), len(listB)
for i in range(len(listA)):
combinedList = listA[i] + listB[i]
newCombinedList = abunHouseKeeper.getDistinct(combinedList)
tmpResolved = resolveConflict(newCombinedList)
resolvedList[i] = tmpResolved
return resolvedList
def filterConfidResolve(resolvedList):
newResolvedList = []
resolvedList.sort()
conThres = abunHouseKeeper.abunGlobalSplitParameterRobot.BRThres
print "conThres", conThres
for key, items in groupby(resolvedList):
tmpList = list(items)
if len(tmpList) >= conThres:
newResolvedList.append(key)
if False:
noConflict = resolveConflict(abunHouseKeeper.getDistinct(resolvedList))
noConflict = abunHouseKeeper.getDistinct(noConflict)
newResolvedList = abunHouseKeeper.getDistinct(newResolvedList)
newResolvedList = abunHouseKeeper.getDistinct(
intersect(newResolvedList, noConflict))
return newResolvedList
def resolveConflictX(listA, listB):
resolvedList = [[] for i in range(len(listA))]
print "len(listA), len(listB)", len(listA), len(listB)
for i in range(len(listA)):
combinedList = listA[i] + listB[i]
newCombinedList = abunHouseKeeper.getDistinct(combinedList)
tmpResolved = resolveConflict(newCombinedList)
resolvedList[i] = tmpResolved
return resolvedList
def findAttachedContigs(rList, side, folderName, sortedReadList, sortedReadDic, lenDicContig,lenDicRead): cList = [] thres = thresMiddleContig for r in rList: if r in sortedReadDic: tmp = sortedReadDic[r] while tmp < len(sortedReadList) and sortedReadList[tmp][-1] == r: eachsub = sortedReadList[tmp] if overlapCRJustREnd(eachsub, side, thres, lenDicContig, lenDicRead): cList.append([eachsub[-2], r]) tmp = tmp + 1 newCList = abunHouseKeeper.getDistinct(cList) return newCList
def findAttachedContigs(rList, side, folderName, sortedReadList, sortedReadDic,
lenDicContig, lenDicRead):
cList = []
thres = thresMiddleContig
for r in rList:
if r in sortedReadDic:
tmp = sortedReadDic[r]
while tmp < len(sortedReadList) and sortedReadList[tmp][-1] == r:
eachsub = sortedReadList[tmp]
if overlapCRJustREnd(eachsub, side, thres, lenDicContig,
lenDicRead):
cList.append([eachsub[-2], r])
tmp = tmp + 1
newCList = abunHouseKeeper.getDistinct(cList)
return newCList
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 abunSplitWithXResolve(folderName, mummerLink, myCountDic, contigReadGraph,
contigFilename, readsetFilename):
N1 = len(myCountDic) * 2
print "N1", N1
# Debug
G = graphLib.seqGraph(0)
G.loadFromFile(folderName, contigReadGraph)
adj = [[] for i in range(N1)]
for i in range(N1):
adj[i] = abunGraphLib.findAllReachable(i, N1, G)
Gnew = graphLib.seqGraph(N1)
for i in range(N1):
for j in adj[i]:
Gnew.insertEdge(i, j, 1)
Gnew.reportEdge()
# End Debug
if False:
json_data = open(folderName + "phaseRepeat.txt", 'r')
repeatPairs = json.load(json_data)
repeatPairs = obtainNonEmpty(repeatPairs)
biResolvedCombineList = []
for eachitem in repeatPairs:
inList, outList = eachitem[0], eachitem[1]
resolvedList = determindMatch(inList, outList, myCountDic,
folderName, contigReadGraph, N1)
biResolvedCombineList += resolvedList
### Xnode repeatResolution
xResolvedList, mapDummyToRealDic = xNodeResolving(
folderName, contigReadGraph)
### Combine resolution
resolvedList = xResolvedList + biResolvedCombineList
resolvedList = abunHouseKeeper.getDistinct(resolvedList)
print "resolvedList, len(resolvedList),len(xResolvedList), len(biResolvedCombineList) ", resolvedList, len(
resolvedList), len(xResolvedList), len(biResolvedCombineList)
with open(folderName + "resolvedList.json", 'w') as f:
json.dump(resolvedList, f)
with open(folderName + "mapDummyToRealDic.json", 'w') as f:
json.dump(mapDummyToRealDic, f)
if False:
json_data = open(folderName + "resolvedList.json", 'r')
resolvedList = json.load(json_data)
json_data = open(folderName + "mapDummyToRealDic.json", 'r')
mapDummyToRealDic = json.load(json_data)
gapContentLookUpList = []
gapContentLookUpList = generateGapContentLookup(
folderName, mummerLink, resolvedList, contigReadGraph,
contigFilename, readsetFilename, mapDummyToRealDic)
gapContentLookUpDic = {}
gapContentLookUpList.sort()
for eachitem in gapContentLookUpList:
gapContentLookUpDic[str(eachitem[0]) + "_" + str(eachitem[1])] = [
eachitem[2], eachitem[3], eachitem[4]
]
print eachitem[2:4], len(eachitem[4])
with open(folderName + "gapContentLookUpDic.json", 'w') as f:
json.dump(gapContentLookUpDic, f)
if False:
json_data = open(folderName + "resolvedList.json", 'r')
resolvedList = json.load(json_data)
json_data = open(folderName + "mapDummyToRealDic.json", 'r')
mapDummyToRealDic = json.load(json_data)
G = graphLib.seqGraph(N1 + len(mapDummyToRealDic))
addEdges(G, resolvedList)
G.condense()
G.saveToFile(folderName, "xResolvedGraph")
if False:
json_data = open(folderName + "mapDummyToRealDic.json", 'r')
mapDummyToRealDic = json.load(json_data)
G = graphLib.seqGraph(0)
G.loadFromFile(folderName, "xResolvedGraph")
json_data = open(folderName + "gapContentLookUpDic.json", 'r')
gapContentLookUpDic = json.load(json_data)
print "Final step: really hacking a file"
os.system("cp " + folderName + contigFilename + "_Double.fasta " +
folderName + "tmpWithDummy.fasta")
contigList = IORobot.readContigsFromFile(
folderName, contigFilename + "_Double.fasta")
f = open(folderName + "tmpWithDummy.fasta", 'a')
for i in range(len(mapDummyToRealDic)):
id = mapDummyToRealDic[str(i)]
f.write(">SegDum" + str(i) + "\n")
f.write(contigList[id] + "\n")
f.close()
IORobot.extractGraphToContigs(G, folderName, mummerLink, "abun.fasta",
"tmpWithDummy.fasta",
gapContentLookUpDic, mapDummyToRealDic)
def identifyRepeat(folderName, mummerLink,contigFilename,contigReadGraph, repeatFilename, optionToRun ):
'''
Input : Graph --- phaseStringGraph1
Output: repeat pairs { [ (1,2), (3,4) ] , [(5,6),(7,8)] }
Algorithm:
a) Reachability test on the graph to find the partners
b) Form Bipartite graph
c) Find connected component in the bipartite and define as repeat pairs
'''
# ## (a) reachability test to find partners
G = graphLib.seqGraph(0)
G.loadFromFile(folderName, contigReadGraph)
# G.reportEdge()
lenDicCC = IORobot.obtainLength(folderName, contigFilename+"_Double.fasta")
adjacencyList = [[] for i in range(len(lenDicCC))]
N1 = len(lenDicCC)
# # Debug
# for i in range(14):
# debugGraphPath(i, 2, G, N1)
# # End Debug
for i in range(len(lenDicCC)):
adjacencyList[i] = abunGraphLib.findAllReachable(i, N1, G)
print "i, adjacencyList[i] : ", i , adjacencyList[i]
# ## (b) formation of bipartite graph
if optionToRun == "tandem" :
newAdjacencyList = adjacencyList
elif optionToRun == "xphase":
newAdjacencyList = abunGraphLib.filterEdge(adjacencyList, folderName, contigFilename)
G2 = abunGraphLib.seqGraphWt(N1 * 2)
for i in range(N1):
for j in newAdjacencyList[i]:
G2.insertEdge(2 * i, 2 * j + 1, 1)
G2.insertEdge(2 * j + 1, 2 * i, 1)
clusters = G2.findConnectedComponents()
repeatList = []
for eachitem in clusters:
leftList, rightList = [], []
for eachsubitem in eachitem:
if eachsubitem % 2 == 0 :
leftList.append(eachsubitem)
else:
rightList.append(eachsubitem)
repeatList.append([abunHouseKeeper.getDistinct(leftList), abunHouseKeeper.getDistinct(rightList)])
with open(folderName + repeatFilename, 'w') as outfile:
json.dump(repeatList, outfile)
json_data = open(folderName + repeatFilename, 'r')
loadData = json.load(json_data)
assert(loadData == repeatList)
def identifyRepeat(folderName, mummerLink, contigFilename, contigReadGraph,
repeatFilename, optionToRun):
'''
Input : Graph --- phaseStringGraph1
Output: repeat pairs { [ (1,2), (3,4) ] , [(5,6),(7,8)] }
Algorithm:
a) Reachability test on the graph to find the partners
b) Form Bipartite graph
c) Find connected component in the bipartite and define as repeat pairs
'''
# ## (a) reachability test to find partners
G = graphLib.seqGraph(0)
G.loadFromFile(folderName, contigReadGraph)
# G.reportEdge()
lenDicCC = IORobot.obtainLength(folderName,
contigFilename + "_Double.fasta")
adjacencyList = [[] for i in range(len(lenDicCC))]
N1 = len(lenDicCC)
# # Debug
# for i in range(14):
# debugGraphPath(i, 2, G, N1)
# # End Debug
for i in range(len(lenDicCC)):
adjacencyList[i] = abunGraphLib.findAllReachable(i, N1, G)
print "i, adjacencyList[i] : ", i, adjacencyList[i]
# ## (b) formation of bipartite graph
if optionToRun == "tandem":
newAdjacencyList = adjacencyList
elif optionToRun == "xphase":
newAdjacencyList = abunGraphLib.filterEdge(adjacencyList, folderName,
contigFilename)
G2 = abunGraphLib.seqGraphWt(N1 * 2)
for i in range(N1):
for j in newAdjacencyList[i]:
G2.insertEdge(2 * i, 2 * j + 1, 1)
G2.insertEdge(2 * j + 1, 2 * i, 1)
clusters = G2.findConnectedComponents()
repeatList = []
for eachitem in clusters:
leftList, rightList = [], []
for eachsubitem in eachitem:
if eachsubitem % 2 == 0:
leftList.append(eachsubitem)
else:
rightList.append(eachsubitem)
repeatList.append([
abunHouseKeeper.getDistinct(leftList),
abunHouseKeeper.getDistinct(rightList)
])
with open(folderName + repeatFilename, 'w') as outfile:
json.dump(repeatList, outfile)
json_data = open(folderName + repeatFilename, 'r')
loadData = json.load(json_data)
assert (loadData == repeatList)
def BResolution(Gnew, folderName, contigReadGraph, N1, myCountDic, lenDic):
if abunHouseKeeper.abunGlobalSplitParameterRobot.runBResolve:
print "abunHouseKeeper.abunGlobalSplitParameterRobot.runBResolve", abunHouseKeeper.abunGlobalSplitParameterRobot.runBResolve
maxRThres = abunHouseKeeper.abunGlobalSplitParameterRobot.RThres
repeatFinder.adjListToRepeatList(Gnew.adj, folderName, "phaseRepeatTR.txt")
json_data = open(folderName + "phaseRepeatTR.txt", "r")
repeatPairs = json.load(json_data)
repeatPairs = obtainNonEmpty(repeatPairs)
biResolvedCombineList = []
G = graphLib.seqGraph(0)
G.loadFromFile(folderName, contigReadGraph)
Grev = abunGraphLib.formReverseGraphFast(G)
abunAnalysisList = []
for eachitem in repeatPairs:
inList, outList = eachitem[0], eachitem[1]
resolvedList, brResolvedList = [], []
if abunHouseKeeper.abunGlobalSplitParameterRobot.toRunAbunB:
if abunHouseKeeper.abunGlobalSplitParameterRobot.AbunLowerB > 0:
abunHouseKeeper.abunGlobalSplitParameterRobot.AbunLower = (
abunHouseKeeper.abunGlobalSplitParameterRobot.AbunLowerB
)
if abunHouseKeeper.abunGlobalSplitParameterRobot.AbunUpperB > 0:
abunHouseKeeper.abunGlobalSplitParameterRobot.AbunUpper = (
abunHouseKeeper.abunGlobalSplitParameterRobot.AbunUpperB
)
if not abunHouseKeeper.abunGlobalSplitParameterRobot.toRunAggB:
resolvedList = determindMatch(inList, outList, myCountDic, folderName, contigReadGraph, N1)
else:
resolvedList = determindMatchAggregate(
inList, outList, myCountDic, folderName, contigReadGraph, N1, Gnew, lenDic
)
if abunHouseKeeper.abunGlobalSplitParameterRobot.toRunBRB:
if abunHouseKeeper.abunGlobalSplitParameterRobot.BRThresB > 0:
abunHouseKeeper.abunGlobalSplitParameterRobot.BRThres = (
abunHouseKeeper.abunGlobalSplitParameterRobot.BRThresB
)
brResolvedList = formBRReolve(folderName, inList, outList, G, Grev, True, N1)
combinedList = abunHouseKeeper.getDistinct(resolvedList + brResolvedList)
print "resolvedList, brResolvedList, inList, outList", resolvedList, brResolvedList, inList, outList
print "resolveConflict(combinedList)", resolveConflict(combinedList)
abunAnalysisList.append([inList, outList, resolvedList, brResolvedList, resolveConflict(combinedList)])
if len(inList) <= maxRThres and len(outList) <= maxRThres and len(inList) > 0 and len(outList) > 0:
# biResolvedCombineList += resolveConflict(combinedList)
resolvedCombine = resolveConflict(combinedList)
### kkdebug
Gnew.bipartiteLocalResolve(resolvedCombine, inList, outList, folderName)
# json_data = open(folderName + "hackBRResolveList.json", 'r')
# dataItem = json.load(json_data)
# Gnew.bipartiteResolve(dataItem)
### end kkdebug
Gnew.condense()
with open(folderName + "biResolvedCombineList.json", "w") as f:
json.dump(biResolvedCombineList, f)
with open(folderName + "abunAnalysisList.json", "w") as f:
json.dump(abunAnalysisList, f)
# assert(1==2)
return Gnew
else:
return Gnew
def abunSplitWithXResolve(folderName, mummerLink, myCountDic, contigReadGraph, contigFilename, readsetFilename):
N1 = len(myCountDic) * 2
print "N1", N1
# Debug
G = graphLib.seqGraph(0)
G.loadFromFile(folderName, contigReadGraph)
adj = [[] for i in range(N1)]
for i in range(N1):
adj[i] = abunGraphLib.findAllReachable(i, N1, G)
Gnew = graphLib.seqGraph(N1)
for i in range(N1):
for j in adj[i]:
Gnew.insertEdge(i, j, 1)
Gnew.reportEdge()
# End Debug
if False:
json_data = open(folderName + "phaseRepeat.txt", "r")
repeatPairs = json.load(json_data)
repeatPairs = obtainNonEmpty(repeatPairs)
biResolvedCombineList = []
for eachitem in repeatPairs:
inList, outList = eachitem[0], eachitem[1]
resolvedList = determindMatch(inList, outList, myCountDic, folderName, contigReadGraph, N1)
biResolvedCombineList += resolvedList
### Xnode repeatResolution
xResolvedList, mapDummyToRealDic = xNodeResolving(folderName, contigReadGraph)
### Combine resolution
resolvedList = xResolvedList + biResolvedCombineList
resolvedList = abunHouseKeeper.getDistinct(resolvedList)
print "resolvedList, len(resolvedList),len(xResolvedList), len(biResolvedCombineList) ", resolvedList, len(
resolvedList
), len(xResolvedList), len(biResolvedCombineList)
with open(folderName + "resolvedList.json", "w") as f:
json.dump(resolvedList, f)
with open(folderName + "mapDummyToRealDic.json", "w") as f:
json.dump(mapDummyToRealDic, f)
if False:
json_data = open(folderName + "resolvedList.json", "r")
resolvedList = json.load(json_data)
json_data = open(folderName + "mapDummyToRealDic.json", "r")
mapDummyToRealDic = json.load(json_data)
gapContentLookUpList = []
gapContentLookUpList = generateGapContentLookup(
folderName, mummerLink, resolvedList, contigReadGraph, contigFilename, readsetFilename, mapDummyToRealDic
)
gapContentLookUpDic = {}
gapContentLookUpList.sort()
for eachitem in gapContentLookUpList:
gapContentLookUpDic[str(eachitem[0]) + "_" + str(eachitem[1])] = [eachitem[2], eachitem[3], eachitem[4]]
print eachitem[2:4], len(eachitem[4])
with open(folderName + "gapContentLookUpDic.json", "w") as f:
json.dump(gapContentLookUpDic, f)
if False:
json_data = open(folderName + "resolvedList.json", "r")
resolvedList = json.load(json_data)
json_data = open(folderName + "mapDummyToRealDic.json", "r")
mapDummyToRealDic = json.load(json_data)
G = graphLib.seqGraph(N1 + len(mapDummyToRealDic))
addEdges(G, resolvedList)
G.condense()
G.saveToFile(folderName, "xResolvedGraph")
if False:
json_data = open(folderName + "mapDummyToRealDic.json", "r")
mapDummyToRealDic = json.load(json_data)
G = graphLib.seqGraph(0)
G.loadFromFile(folderName, "xResolvedGraph")
json_data = open(folderName + "gapContentLookUpDic.json", "r")
gapContentLookUpDic = json.load(json_data)
print "Final step: really hacking a file"
os.system("cp " + folderName + contigFilename + "_Double.fasta " + folderName + "tmpWithDummy.fasta")
contigList = IORobot.readContigsFromFile(folderName, contigFilename + "_Double.fasta")
f = open(folderName + "tmpWithDummy.fasta", "a")
for i in range(len(mapDummyToRealDic)):
id = mapDummyToRealDic[str(i)]
f.write(">SegDum" + str(i) + "\n")
f.write(contigList[id] + "\n")
f.close()
IORobot.extractGraphToContigs(
G, folderName, mummerLink, "abun.fasta", "tmpWithDummy.fasta", gapContentLookUpDic, mapDummyToRealDic
)
def formConfirmReadResolve(folderName, inList, outList, G, Grev, N1):
# print "formConfirmReadResolve"
resolvedList = []
confirmingReadList = []
brLFlankList = []
brRFlankList = []
### Find possible candidate reads
print "inList , outList formConfirmReadResolve()", inList, outList
for eachin in inList:
for eachout in outList:
pathList = abunGraphLib.findAllPathK(eachin, eachout, G, 3)
for path in pathList:
if len(path) == 3 and path[1] >= N1:
R = path[1]
confirmingReadList.append(R)
brLFlankList.append([eachin, R])
brRFlankList.append([eachout, R])
### Filter simple false cases
toUseReadDic = {}
confirmingReadList.sort()
for key, items in groupby(confirmingReadList):
toUseReadDic[str(key)] = True
newbrLFlankList = abunHouseKeeper.getDistinct(brLFlankList)
newbrLFlankList.sort(key=itemgetter(1))
for key, items in groupby(newbrLFlankList, itemgetter(1)):
mylist = list(items)
if len(mylist) > 1:
toUseReadDic[str(key)] = False
newbrRFlankList = abunHouseKeeper.getDistinct(brRFlankList)
newbrRFlankList.sort(key=itemgetter(1))
for key, items in groupby(newbrRFlankList, itemgetter(1)):
mylist = list(items)
if len(mylist) > 1:
toUseReadDic[str(key)] = False
finalSearchReadList = []
for eachitem in toUseReadDic:
if toUseReadDic[eachitem] == True:
finalSearchReadList.append(int(eachitem))
### Check paths to confirm all false cases
for eachR in finalSearchReadList:
l1 = abunGraphLib.findAllReachable(eachR, N1, G)
l2 = abunGraphLib.findAllReachable(eachR, N1, Grev)
l1Distinct = abunHouseKeeper.getDistinct(l1)
l2Distinct = abunHouseKeeper.getDistinct(l2)
if len(l1Distinct) == 1 and len(l2Distinct) == 1:
c1, c2 = l1Distinct[0], l2Distinct[0]
resolvedList.append([c2, c1])
return resolvedList
def BResolution(Gnew, folderName, contigReadGraph, N1, myCountDic, lenDic,
mummerLink):
if abunHouseKeeper.abunGlobalSplitParameterRobot.runBResolve:
print "abunHouseKeeper.abunGlobalSplitParameterRobot.runBResolve", abunHouseKeeper.abunGlobalSplitParameterRobot.runBResolve
maxRThres = abunHouseKeeper.abunGlobalSplitParameterRobot.RThres
repeatFinder.adjListToRepeatList(Gnew.adj, folderName,
"phaseRepeatTR.txt")
json_data = open(folderName + "phaseRepeatTR.txt", 'r')
repeatPairs = json.load(json_data)
repeatPairs = obtainNonEmpty(repeatPairs)
biResolvedCombineList = []
G = abunGraphLib.seqGraphWt(0)
G.loadFromFile(folderName, contigReadGraph)
Grev = abunGraphLib.formReverseGraphFast(G)
abunAnalysisList = []
for eachitem in repeatPairs:
inList, outList = eachitem[0], eachitem[1]
if not abunHouseKeeper.abunGlobalRunEM:
resolvedList, brResolvedList = [], []
if abunHouseKeeper.abunGlobalSplitParameterRobot.toRunAbunB:
if abunHouseKeeper.abunGlobalSplitParameterRobot.AbunLowerB > 0:
abunHouseKeeper.abunGlobalSplitParameterRobot.AbunLower = abunHouseKeeper.abunGlobalSplitParameterRobot.AbunLowerB
if abunHouseKeeper.abunGlobalSplitParameterRobot.AbunUpperB > 0:
abunHouseKeeper.abunGlobalSplitParameterRobot.AbunUpper = abunHouseKeeper.abunGlobalSplitParameterRobot.AbunUpperB
if not abunHouseKeeper.abunGlobalSplitParameterRobot.toRunAggB:
resolvedList = determindMatch(inList, outList,
myCountDic, folderName,
contigReadGraph, N1)
else:
resolvedList = determindMatchAggregate(
inList, outList, myCountDic, folderName,
contigReadGraph, N1, Gnew, lenDic)
if abunHouseKeeper.abunGlobalSplitParameterRobot.toRunBRB:
if abunHouseKeeper.abunGlobalSplitParameterRobot.BRThresB > 0:
abunHouseKeeper.abunGlobalSplitParameterRobot.BRThres = abunHouseKeeper.abunGlobalSplitParameterRobot.BRThresB
brResolvedList = formBRReolve(folderName, inList, outList,
G, Grev, True, N1)
combinedList = abunHouseKeeper.getDistinct(resolvedList +
brResolvedList)
print "resolvedList, brResolvedList, inList, outList", resolvedList, brResolvedList, inList, outList
print "resolveConflict(combinedList)", resolveConflict(
combinedList)
abunAnalysisList.append([
inList, outList, resolvedList, brResolvedList,
resolveConflict(combinedList)
])
if len(inList) <= maxRThres and len(
outList) <= maxRThres and len(inList) > 0 and len(
outList) > 0:
resolvedCombine = resolveConflict(combinedList)
Gnew.bipartiteLocalResolve(resolvedCombine, inList,
outList, folderName)
else:
import emalgo
resolvedCombine = emalgo.BResolvePreparation(
folderName, inList, outList, G, Grev, N1, mummerLink)
Gnew.bipartiteLocalResolve(resolvedCombine, inList, outList,
folderName)
Gnew.condense()
with open(folderName + "biResolvedCombineList.json", 'w') as f:
json.dump(biResolvedCombineList, f)
with open(folderName + "abunAnalysisList.json", 'w') as f:
json.dump(abunAnalysisList, f)
#assert(1==2)
return Gnew
else:
return Gnew
