def bitScoreMM(pwmFileName,
genomeDict,
mpbsDict,
scoringMethod,
tempLocation,
pseudocounts=0.1,
bitscore=12.0,
fpr=0.01,
precision=10**4,
highCutoff=0.7,
functionalDepth=0.9):
"""Performs basic motif matching algorithm and writes the results to a dictionary indexed by chromosome.
Keyword arguments:
pwmFileName -- PWM file name.
genomeDict -- Genome dictionary.
mpbsDict -- Dictionary of MPBSs to insert the results.
scoringMethod -- Method to evaluate which MPBSs are enriched.
tempLocation -- Location to write temporary PWM files in order to help PWM creation and pseudocounting.
pseudocounts -- Amount of pseudocounts to add in each PWM matrix's cell. (default 0.1)
bitscore -- The cutoff bitscore value. (default 12.0)
fpr -- False positive rate to determine the cutoff value. (default 0.01)
precision -- Motif score distribution precision. (default 10**4)
highCutoff -- High cutoff for Boyle's rule. (default 0.7)
functionalDepth -- Functional depth for Boyle's rule. (default 0.9)
Returns:
mpbsDict -- This method inserts entries on the mpbsDict.
"""
# Reading PWM
pwm = createPwmDict(pwmFileName, pseudocounts)
pwmName = pwmFileName.split("/")[-1].split(".")[0]
pwmLen = len(pwm["A"])
background = math.log(0.25, 2) * pwmLen
# Evaluating threshold
pwmThreshold = 0.0
if (scoringMethod == "bitscore"):
pwmThreshold = bitscore
elif (scoringMethod == "fpr"):
bioPwm = biopythonMM.readPwmFile(pwmFileName, tempLocation,
pseudocounts)
sd = Motif.ScoreDistribution(bioPwm, precision=precision)
pwmThreshold = sd.threshold_fpr(fpr)
elif (scoringMethod == "boyle"):
maxScore = 0.0
minScore = 0.0 # TODO Boyle's rule is not suited for negative values.
for i in range(0, pwmLen):
maxScore += max(pwm["A"][i], pwm["C"][i], pwm["G"][i], pwm["T"][i])
maxScore -= background
pwmThreshold = min(highCutoff * maxScore,
functionalDepth * (maxScore - minScore))
else:
sys.stderr.write("Choose a valid scoring method.\n")
sys.exit(0)
# Creating aditional parameters
chrList = constants.getChromList(reference=[mpbsDict])
tempMpbsDict = dict([(e, []) for e in chrList])
maxValue = -99.0
revDict = dict([("A", "T"), ("T", "A"), ("C", "G"), ("G", "C"),
("N", "N")])
# Iterating on chromosomes
for chrName in chrList:
# Reading genome
sequence = genomeDict[chrName].upper()
# Performing motif matching
for pos in xrange(0, len(sequence) - pwmLen + 1):
scoreF = -background
scoreR = -background
for i in range(0, pwmLen):
scoreF += pwm[sequence[pos + i]][i]
scoreR += pwm[revDict[sequence[pos + pwmLen - i - 1]]][i]
if (scoreF > pwmThreshold):
if (scoreF > maxValue): maxValue = scoreF
tempMpbsDict[chrName].append(
[pos, pos + pwmLen, pwmName, scoreF, "+"])
if (scoreR > pwmThreshold):
if (scoreR > maxValue): maxValue = scoreR
tempMpbsDict[chrName].append(
[pos, pos + pwmLen, pwmName, scoreR, "-"])
# Update scores - new scores are within [0,1000]
for chrName in chrList:
for e in tempMpbsDict[chrName]:
mpbsDict[chrName].append([
e[0], e[1], e[2],
int(1000 * (e[3] - pwmThreshold) / (maxValue - pwmThreshold)),
e[4]
])
return 0
def biopythonMM(pwmFileName,
genomeDict,
mpbsDict,
scoringMethod,
tempLocation,
pseudocounts=0.1,
bitscore=12.0,
fpr=0.01,
precision=10**4,
highCutoff=0.7,
functionalDepth=0.9):
"""Performs Biopython based motif matching and writes the results to a dictionary indexed by chromosome.
Keyword arguments:
pwmFileName -- PWM file name.
genomeDict -- Genome dictionary.
mpbsDict -- Dictionary of MPBSs to insert the results.
scoringMethod -- Method to evaluate which MPBSs are enriched.
tempLocation -- Location to write temporary PWM files in order to help PWM creation and pseudocounting.
pseudocounts -- Amount of pseudocounts to add in each PWM matrix's cell. (default 0.1)
bitscore -- The cutoff bitscore value. (default 12.0)
fpr -- False positive rate to determine the cutoff value. (default 0.01)
precision -- Motif score distribution precision. (default 10**4)
highCutoff -- High cutoff for Boyle's rule. (default 0.7)
functionalDepth -- Functional depth for Boyle's rule. (default 0.9)
Returns:
mpbsDict -- This method inserts entries on the mpbsDict.
"""
# Reading PWM
pwm = readPwmFile(pwmFileName, tempLocation, pseudocounts)
pwmName = pwmFileName.split("/")[-1].split(".")[0]
pwmLen = len(pwm)
# Evaluating threshold
pwmThreshold = 0.0
if (scoringMethod == "bitscore"):
pwmThreshold = bitscore
elif (scoringMethod == "fpr"):
sd = Motif.ScoreDistribution(pwm, precision=precision)
pwmThreshold = sd.threshold_fpr(fpr)
elif (scoringMethod == "boyle"):
maxScore = pwm.max_score()
minScore = 0.0 # TODO Boyle's rule is not suited for negative values.
pwmThreshold = min(highCutoff * maxScore,
functionalDepth * (maxScore - minScore))
else:
sys.stderr.write("Choose a valid scoring method.\n")
sys.exit(0)
# Creating aditional parameters
chrList = constants.getChromList(reference=[mpbsDict])
tempMpbsDict = dict([(e, []) for e in chrList])
maxValue = -99.0
# Iterating on chromosomes
for chrName in chrList:
# Reading genome
sequence = genomeDict[chrName]
# Performing biopython's motif matching
for pos, score in pwm.search_pwm(sequence, threshold=pwmThreshold):
if (score > maxValue): maxValue = score
if (pos >= 0):
tempMpbsDict[chrName].append(
[pos, pos + pwmLen, pwmName, score, "+"])
else:
tempMpbsDict[chrName].append(
[-pos, -pos + pwmLen, pwmName, score, "-"])
# Update scores - new scores are within [0,1000]
for chrName in chrList:
for e in tempMpbsDict[chrName]:
mpbsDict[chrName].append([
e[0], e[1], e[2],
int(1000 * (e[3] - pwmThreshold) / (maxValue - pwmThreshold)),
e[4]
])
return 0
def motifMatchingBiopython(combinationList,pwmList,coordDict,pwmLocation,genomeList,tempLocation,fpr=0.01,pseudocounts=0.0,precision=10**4,color="black"):
"""Performs Biopython based motif matching and returns a list containing the matches and
writes the results on bed files.
Keyword arguments:
combinationList -- List of the number of cobinding combinations.
pwmList -- List of PWMs where each entry represents the name of a PWM file.
coordDict -- Dictionary of coordinates where the motif matching will be applied.
pwmLocation -- Path containing the motif pwm files.
genomeList -- List of fasta files containing the sequences to perform the motif matching, where the headers are the chromosomes.
tempLocation -- Location to write temporary PWM files in order to help PWM creation and pseudocounting.
fpr -- False positive rate to determine the cutoff value. (default 0.01)
pseudocounts -- Amount of pseudocounts to add in each PWM matrix's cell. (default 0.0)
precision -- Motif score distribution precision. (default 10**4)
color -- Color of the bed entries. Can be 'green', 'red' or 'black'. (default 'black')
Returns:
mpbsDict -- Dictionary (for each PWM) of dictionaries (for each chromosome) of motif predicted binding sites.
statDict -- Dictionary of statistics for Fisher test concerning the number of motifs inside enriched regions.
geneDict -- Dictionary of genes (position NAME in bed file) that contains each motif.
"""
# Reading PWM
pwmDict = dict()
for pwmName in pwmList: pwmDict[pwmName] = readPwmFile(pwmLocation+pwmName+".pwm","/".join(tempLocation.split("/")[:-1])+"/",pseudocounts)
# Evaluating thresholds
pwmThresholdDict = dict()
for pwmName in pwmList:
sd = Motif.ScoreDistribution(pwmDict[pwmName],precision=precision)
pwmThresholdDict[pwmName] = sd.threshold_fpr(fpr)
# Reading genome
genomeDict = genome.readFastaFiles(genomeList)
# Creating chromosome list
chrList = constants.getChromList(reference=[coordDict])
# Removing chrX, chrY and chrM
# TODO Stop removing these chromosomes
#chrListT = []
#for e in chrList:
# if(e not in ["chrX", "chrY", "chrM"]): chrListT.append(e)
#chrList = chrListT
# Evaluating bed additionals
if(color == "green"): color = "0,130,0"
elif(color == "red"): color = "130,0,0"
elif(color == "black"): color = "0,0,0"
# Create combinations dictionary keys
combKeys = []
for c in combinationList:
for b in [",".join(e) for e in itertools.combinations(pwmList,c)]: combKeys.append(b)
# Iterating on chromosomes
mpbsDict = dict([(e,dict()) for e in pwmDict.keys()])
statDict = dict([(e,[0,0]) for e in combKeys]) # Left is evidence / Right is not evidence
geneDict = dict([(e,[]) for e in combKeys])
maxDict = dict([(e,-99.0) for e in pwmDict.keys()])
ct=0
for chrName in chrList:
# Reading genome
if(chrName not in genomeDict.keys()): continue
sequence = genomeDict[chrName]
# Iterating on coordinate dictionary
for e in mpbsDict.keys(): mpbsDict[e][chrName] = []
for coord in coordDict[chrName]:
ct=ct+1
#print "region", ct
# Getting current sequence based on coordinates
currSeq = sequence[coord[0]:coord[1]]
# Keeping track of the factors found in this coordinate
flagMotifs = dict([(e,False) for e in pwmDict.keys()])
# Iterating on PWMs
for pwmName in pwmDict.keys():
pwmLen = len(pwmDict[pwmName])
for pos, score in pwmDict[pwmName].search_pwm(currSeq,threshold=pwmThresholdDict[pwmName]):
if(score > maxDict[pwmName]): maxDict[pwmName] = score
if(pos >= 0): mpbsDict[pwmName][chrName].append([pos+coord[0],pos+coord[0]+pwmLen,pwmName,score,"+",pos+coord[0],pos+coord[0]+pwmLen,color])
else: mpbsDict[pwmName][chrName].append([-pos+coord[0],-pos+coord[0]+pwmLen,pwmName,score,"-",-pos+coord[0],-pos+coord[0]+pwmLen,color])
flagMotifs[pwmName] = True
# Updating statistic counts and genes
motifsFoundList = [k for k in pwmList if flagMotifs[k]]
motifsFoundKeys = []
motifsNotFoundKeys = [e for e in combKeys]
for c in combinationList:
for b in [",".join(e) for e in itertools.combinations(motifsFoundList,c)]:
motifsFoundKeys.append(b)
motifsNotFoundKeys.remove(b)
for k in motifsFoundKeys:
statDict[k][0] += 1
for e in coord[2].split(":"): geneDict[k].append(e)
for k in motifsNotFoundKeys:
statDict[k][1] += 1
# Update scores - new scores are within [0,1000]
for pwmName in pwmDict.keys():
for chrName in mpbsDict[pwmName].keys():
for e in mpbsDict[pwmName][chrName]:
e[3] = int(1000*(e[3]-pwmThresholdDict[pwmName])/(maxDict[pwmName]-pwmThresholdDict[pwmName]))
# Remove repetitive genes from geneList
for k in geneDict.keys(): geneDict[k] = list(set(geneDict[k]))
return mpbsDict, statDict, geneDict
def fimoMM(pwmFileName,
genomeFile,
mpbsDict,
scoringMethod,
tempLocation,
pseudocounts=0.1,
bitscore=12.0,
fpr=0.01,
precision=10**4,
highCutoff=0.7,
functionalDepth=0.9,
threshold=0.0001):
"""Performs FIMO motif matching algorithm and writes the results to a dictionary indexed by chromosome.
Keyword arguments:
pwmFileName -- PWM file name.
genomeFile -- Fasta file containing the regions to be analyzed
mpbsDict -- Dictionary of MPBSs to insert the results.
scoringMethod -- Method to evaluate which MPBSs are enriched.
tempLocation -- Location to write temporary PWM files in order to help PWM creation and pseudocounting.
pseudocounts -- Amount of pseudocounts to add in each PWM matrix's cell. (default 0.1)
bitscore -- The cutoff bitscore value. (default 12.0)
fpr -- False positive rate to determine the cutoff value. (default 0.01)
precision -- Motif score distribution precision. (default 10**4)
highCutoff -- High cutoff for Boyle's rule. (default 0.7)
functionalDepth -- Functional depth for Boyle's rule. (default 0.9)
threshold -- The cutoff threshold value. (default 0.0001)
Returns:
mpbsDict -- This method inserts entries on the mpbsDict.
"""
# Converting jaspar to MEME
memeFileName = jasparToMeme(pwmFileName, tempLocation, pseudocounts)
tempPath = "/".join(memeFileName.split("/")[:-1]) + "/"
fimoFileName = tempPath + "results.txt"
errorOutputName = tempPath + "error.txt"
# Evaluating threshold
pwmThreshold = 0.0
if (scoringMethod == "bitscore"):
pwmThreshold = bitscore
threshold = 0.1
elif (scoringMethod == "fpr"):
bioPwm = biopythonMM.readPwmFile(pwmFileName, tempLocation,
pseudocounts)
sd = Motif.ScoreDistribution(bioPwm, precision=precision)
pwmThreshold = sd.threshold_fpr(fpr)
threshold = 0.1
print bioPwm.max_score()
elif (scoringMethod == "boyle"):
maxScore = 0.0
minScore = 0.0 # TODO Boyle's rule is not suited for negative values.
pwmBoyle = bitScoreMM.createPwmDict(pwmFileName, pseudocounts)
pwmLen = len(pwmBoyle["A"])
for i in range(0, pwmLen):
maxScore += max(pwmBoyle["A"][i], pwmBoyle["C"][i],
pwmBoyle["G"][i], pwmBoyle["T"][i])
background = math.log(0.25, 2) * pwmLen
maxScore -= background
pwmThreshold = min(highCutoff * maxScore,
functionalDepth * (maxScore - minScore))
threshold = 0.1
elif (scoringMethod == "fimo"):
pass
else:
sys.stderr.write("Choose a valid scoring method.\n")
sys.exit(0)
# Performing FIMO
os.system(
"fimo --text --verbosity 1 --max-stored-scores 1000000 --output-pthresh "
+ str(threshold) + " " + memeFileName + " " + genomeFile + " > " +
fimoFileName + " 2> " + errorOutputName)
# Reading FIMO output
tempMpbsDict = dict()
fimoFile = open(fimoFileName, "r")
fimoFile.readline()
maxValue = -999
for line in fimoFile:
ll = line.strip().split("\t")
ll = [ll[0][0], ll[0][1:]] + ll[1:]
if (scoringMethod != "fimo" and float(ll[5]) < pwmThreshold): continue
if (float(ll[5]) > maxValue): maxValue = float(ll[5])
if (ll[2] in tempMpbsDict.keys()):
if (ll[0] == "+"):
tempMpbsDict[ll[2]].append(
[int(ll[3]) - 1,
int(ll[4]), ll[1],
float(ll[5]), ll[0]])
else:
tempMpbsDict[ll[2]].append(
[int(ll[4]) - 1,
int(ll[3]), ll[1],
float(ll[5]), ll[0]])
else:
if (ll[0] == "+"):
tempMpbsDict[ll[2]] = [[
int(ll[3]) - 1,
int(ll[4]), ll[1],
float(ll[5]), ll[0]
]]
else:
tempMpbsDict[ll[2]] = [[
int(ll[4]) - 1,
int(ll[3]), ll[1],
float(ll[5]), ll[0]
]]
fimoFile.close()
# Update scores and remove MPBSs with score below pwmThreshold (if it is being used)
for chrName in tempMpbsDict.keys():
for e in tempMpbsDict[chrName]:
if (chrName in mpbsDict.keys()):
mpbsDict[chrName].append([
e[0], e[1], e[2],
int(1000 * (e[3] - pwmThreshold) /
(maxValue - pwmThreshold)), e[4]
])
else:
mpbsDict[chrName] = [[
e[0], e[1], e[2],
int(1000 * (e[3] - pwmThreshold) /
(maxValue - pwmThreshold)), e[4]
]]
# Removing temporary PWM folder
os.system("rm -rf " + "/".join(memeFileName.split("/")[:-1]))
return 0
