def numShifts(rna, secStr): #brute force computation of number shifts
#rna and secStr is 1-indexed
if VERBOSE_PRINT:
print "Sec str: %s" % secStr
n = len(rna)
rna0 = rna
secStr0 = secStr
rna = '$' + rna
secStr = '$' + secStr
SS = basePairList(secStr)
num = 0
tempSS = copy.deepcopy(SS)
for (i, j) in SS:
tempSS.remove((i, j))
if VERBOSE_PRINT:
print "Remove (%d,%d) from %s" % (i, j, secStr[1:])
for x in range(1, n + 1):
if abs(i - x) > THETA and x != j and basePair(rna[i], rna[x]):
if i < x:
tempSS.append((i, x))
if isSecStr(tempSS):
num += 1
if VERBOSE_PRINT:
ss = basePairList2dotBracketNotation(rna0, tempSS)
print '(%d,%d)\t%s' % (i, x, ss)
tempSS.remove((i, x))
else: #i>x
tempSS.append((x, i))
if isSecStr(tempSS):
num += 1
if VERBOSE_PRINT:
ss = basePairList2dotBracketNotation(rna0, tempSS)
print '(%d,%d)\t%s' % (x, i, ss)
tempSS.remove((x, i))
elif abs(j - x) > THETA and x != i and basePair(rna[j], rna[x]):
if j < x:
tempSS.append((j, x))
if isSecStr(tempSS):
num += 1
if VERBOSE_PRINT:
ss = basePairList2dotBracketNotation(rna0, tempSS)
print '(%d,%d)\t%s' % (j, x, ss)
tempSS.remove((j, x))
else: #j>x
tempSS.append((x, j))
if isSecStr(tempSS):
num += 1
if VERBOSE_PRINT:
ss = basePairList2dotBracketNotation(rna0, tempSS)
print '(%d,%d)\t%s' % (x, j, ss)
tempSS.remove((x, j))
tempSS.append((i, j)) #put back the base pair temporarily removed
if PRINT:
print "%s has %s shifts" % (secStr[1:], num
) #recall '$' prepended to secStr
return num
def numShifts(rna,secStr): #brute force computation of number shifts
#rna and secStr is 1-indexed
if VERBOSE_PRINT:
print "Sec str: %s" % secStr
n = len(rna)
rna0 = rna
secStr0= secStr
rna = '$'+rna
secStr = '$'+secStr
SS = basePairList(secStr)
num = 0; tempSS = copy.deepcopy(SS)
for (i,j) in SS:
tempSS.remove( (i,j) )
if VERBOSE_PRINT:
print "Remove (%d,%d) from %s" % (i,j,secStr[1:])
for x in range(1,n+1):
if abs(i-x)>THETA and x!=j and basePair(rna[i],rna[x]):
if i<x:
tempSS.append( (i,x) )
if isSecStr(tempSS):
num += 1
if VERBOSE_PRINT:
ss = basePairList2dotBracketNotation(rna0,tempSS)
print '(%d,%d)\t%s' % (i,x,ss)
tempSS.remove( (i,x) )
else: #i>x
tempSS.append( (x,i) )
if isSecStr(tempSS):
num += 1
if VERBOSE_PRINT:
ss = basePairList2dotBracketNotation(rna0,tempSS)
print '(%d,%d)\t%s' % (x,i,ss)
tempSS.remove( (x,i) )
elif abs(j-x)>THETA and x!= i and basePair(rna[j],rna[x]):
if j<x:
tempSS.append( (j,x) )
if isSecStr(tempSS):
num += 1
if VERBOSE_PRINT:
ss = basePairList2dotBracketNotation(rna0,tempSS)
print '(%d,%d)\t%s' % (j,x,ss)
tempSS.remove( (j,x) )
else: #j>x
tempSS.append( (x,j) )
if isSecStr(tempSS):
num += 1
if VERBOSE_PRINT:
ss = basePairList2dotBracketNotation(rna0,tempSS)
print '(%d,%d)\t%s' % (x,j,ss)
tempSS.remove( (x,j) )
tempSS.append( (i,j) ) #put back the base pair temporarily removed
if PRINT:
print "%s has %s shifts" % (secStr[1:],num) #recall '$' prepended to secStr
return num
def get2DcontactOrder(secStr):
co=0.0
bpList = misc.basePairList(secStr)
for bp in bpList:
co += abs(bp[1]-bp[0])
if len(bpList) !=0:
a_co = float(co)/len(bpList)
r_co = co/(len(secStr)*len(bpList))
else:
a_co = float("inf")
r_co = float("inf")
return a_co,r_co
def get2DcontactOrder(secStr):
co = 0.0
bpList = misc.basePairList(secStr)
for bp in bpList:
co += abs(bp[1] - bp[0])
if len(bpList) != 0:
a_co = float(co) / len(bpList)
r_co = co / (len(secStr) * len(bpList))
else:
a_co = float("inf")
r_co = float("inf")
return a_co, r_co
def numBasePairAdditionsRemovals(rna,secStr): #brute force
#rna and secStr is 1-indexed
if PRINTbasepairadditionsremovals:
print "Sec str: %s" % secStr
n = len(rna)
rna0 = rna
rna = '$'+rna
secStr = '$'+secStr
SS = basePairList(secStr)
num = len(SS); tempSS = copy.deepcopy(SS)
for x in range(1,n-THETA):
for y in range(x+THETA+1,n+1):
if basePair(rna[x],rna[y]):
if (x,y) not in tempSS:
tempSS.append( (x,y) )
if isSecStr(tempSS):
num += 1
if PRINTbasepairadditionsremovals:
ss = basePairList2dotBracketNotation(rna0,tempSS)
print '(%d,%d)\t%s' % (x,y,ss)
tempSS.remove( (x,y) )
return num
def numBasePairAdditionsRemovals(rna, secStr): #brute force
#rna and secStr is 1-indexed
if PRINTbasepairadditionsremovals:
print "Sec str: %s" % secStr
n = len(rna)
rna0 = rna
rna = '$' + rna
secStr = '$' + secStr
SS = basePairList(secStr)
num = len(SS)
tempSS = copy.deepcopy(SS)
for x in range(1, n - THETA):
for y in range(x + THETA + 1, n + 1):
if basePair(rna[x], rna[y]):
if (x, y) not in tempSS:
tempSS.append((x, y))
if isSecStr(tempSS):
num += 1
if PRINTbasepairadditionsremovals:
ss = basePairList2dotBracketNotation(rna0, tempSS)
print '(%d,%d)\t%s' % (x, y, ss)
tempSS.remove((x, y))
return num
def main(filename):
file = open(filename)
rna0 = file.readline().strip()
secStr0 = computeViennaSecStr(rna0)
D = {}
SS = {}
H = {} #H is entropy
n = len(rna0)
num = 0
line = file.readline()
numMut = 0
numSamples = 0
while line:
if line[0] == '>':
if numMut == 0: #first time, get number of samples
numSamples = int(line.split()[2])
else: #update entropy H
H[numMut] = {}
for i in range(1, n + 1): #indices 1<=i<=n
sumEntropyForI = 0.0 #compute H[i]
probII = 1.0 #compute p(i,i) by 1.0 - p(i,j) all j
for j in range(1, n + 1):
if i != j:
probIJ = SS[numMut][i][j] / float(numSamples)
sumEntropyForI += -xLogX(probIJ)
probII -= probIJ
H[numMut][i] = sumEntropyForI + -xLogX(probII)
#Now add information for new number of mutations
words = line.split()
numMut = int(words[-2]) #new number of mutations
D[numMut] = {}
SS[numMut] = {}
for i in range(1, n + 1):
SS[numMut][i] = {}
for j in range(1, n + 1):
SS[numMut][i][j] = 0.0
D[numMut][i - 1] = 0.0
#D[numMut][i] is number of mutations in position i
#WARNING: indices in D are 0<=i<n, while those in SS are 1<=i<=n
num = 0 #start over counter of number of mutations
line = file.readline()
continue
num += 1
if DEBUG: print numMut, num
rna = line.strip().upper()
secStr = file.readline().strip()
bps = basePairList(secStr)
for i in range(n):
if rna0[i] != rna[i]: D[numMut][i] += 1.0
baseI = i + 1 #warning 1<=baseI<=n, but 0<=i<n for accessing RNA string
for j in range(1, n + 1):
if baseI < j and (baseI, j) in bps:
SS[numMut][baseI][j] += 1
elif j < baseI and (j, baseI) in bps:
SS[numMut][baseI][j] += 1
line = file.readline()
file.close()
#Must complete computation of H for last value of numMut
#As well, we normalize D values to be between [0,1]
H[numMut] = {}
for i in range(1, n + 1): #indices 1<=i<=n
sumEntropyForI = 0.0 #compute H[i]
probII = 1.0 #compute p(i,i) by 1.0 - p(i,j) all j
for j in range(1, n + 1):
if i != j:
probIJ = SS[numMut][i][j] / float(num)
sumEntropyForI += -xLogX(probIJ)
probII -= probIJ
H[numMut][i] = sumEntropyForI + -xLogX(probII)
#Normalize D values
for num in range(1, numMut + 1):
for i in range(n):
D[num][i] /= numSamples
#Now compute correlation coefficient
corrCoeffList = []
print "k\tcorrCoeff\tmean1\t\tstdev1\t\tmean2\t\tstdev2"
for num in range(1, numMut + 1):
L1 = []
L2 = []
for i in range(n):
L1.append(D[num][i])
L2.append(H[num][i + 1]) #WARNING: indices in H are from 1 to n
mean1, stdev1, max1, min1 = getSampleStats(L1)
mean2, stdev2, max2, min2 = getSampleStats(L2)
print "%d\t%f\t%f\t%f\t%f\t%f" % (num, corrCoeff(
L1, L2), mean1, stdev1, mean2, stdev2)
def main(filename):
file = open(filename)
rna0 = file.readline().strip()
secStr0 = computeViennaSecStr(rna0)
D = {}; SS = {}; H = {} #H is entropy
n = len(rna0); num = 0
line = file.readline()
numMut = 0; numSamples = 0
while line:
if line[0]=='>':
if numMut == 0: #first time, get number of samples
numSamples = int(line.split()[2])
else: #update entropy H
H[numMut] = {}
for i in range(1,n+1): #indices 1<=i<=n
sumEntropyForI = 0.0 #compute H[i]
probII = 1.0 #compute p(i,i) by 1.0 - p(i,j) all j
for j in range(1,n+1):
if i!=j:
probIJ = SS[numMut][i][j]/float(numSamples)
sumEntropyForI += -xLogX(probIJ)
probII -= probIJ
H[numMut][i] = sumEntropyForI + -xLogX(probII)
#Now add information for new number of mutations
words = line.split()
numMut = int(words[-2]) #new number of mutations
D[numMut] = {}
SS[numMut] = {}
for i in range(1,n+1):
SS[numMut][i] = {}
for j in range(1,n+1): SS[numMut][i][j] = 0.0
D[numMut][i-1] = 0.0
#D[numMut][i] is number of mutations in position i
#WARNING: indices in D are 0<=i<n, while those in SS are 1<=i<=n
num = 0 #start over counter of number of mutations
line = file.readline()
continue
num += 1
if DEBUG: print numMut,num
rna = line.strip().upper()
secStr = file.readline().strip()
bps = basePairList(secStr)
for i in range(n):
if rna0[i]!=rna[i]: D[numMut][i] += 1.0
baseI = i+1 #warning 1<=baseI<=n, but 0<=i<n for accessing RNA string
for j in range(1,n+1):
if baseI<j and (baseI,j) in bps:
SS[numMut][baseI][j] += 1
elif j<baseI and (j,baseI) in bps:
SS[numMut][baseI][j] += 1
line = file.readline()
file.close()
#Must complete computation of H for last value of numMut
#As well, we normalize D values to be between [0,1]
H[numMut] = {}
for i in range(1,n+1): #indices 1<=i<=n
sumEntropyForI = 0.0 #compute H[i]
probII = 1.0 #compute p(i,i) by 1.0 - p(i,j) all j
for j in range(1,n+1):
if i!=j:
probIJ = SS[numMut][i][j]/float(num)
sumEntropyForI += -xLogX(probIJ)
probII -= probIJ
H[numMut][i] = sumEntropyForI + -xLogX(probII)
#Normalize D values
for num in range(1,numMut+1):
for i in range(n): D[num][i] /= numSamples
#Now compute correlation coefficient
corrCoeffList = []
print "k\tcorrCoeff\tmean1\t\tstdev1\t\tmean2\t\tstdev2"
for num in range(1,numMut+1):
L1 = []; L2 = []
for i in range(n):
L1.append(D[num][i])
L2.append(H[num][i+1]) #WARNING: indices in H are from 1 to n
mean1,stdev1,max1,min1 = getSampleStats(L1)
mean2,stdev2,max2,min2 = getSampleStats(L2)
print "%d\t%f\t%f\t%f\t%f\t%f" % (num,corrCoeff(L1,L2),mean1,stdev1,mean2,stdev2)