def consensusMultipleAlign(seqs, threshold, simMatrix):
n = len(seqs)
multipleAlign = []
i = 0
for j in range(i + 1, n):
seqB = seqs[j]
if not multipleAlign:
seqA = seqs[i]
score, alignA, alignB = sequenceAlign(seqA, seqB, simMatrix)
multipleAlign.append(alignA)
multipleAlign.append(alignB)
else:
seqA = consensus(multipleAlign, threshold)
score, alignA, alignB = sequenceAlign(seqA, seqB, simMatrix)
gaps = []
for k, letter in enumerate(alignA):
if letter == '-':
gaps.append(k)
for k, seq in enumerate(multipleAlign):
for gap in gaps:
seq = seq[:gap] + '-' + seq[gap:]
multipleAlign[k] = seq
multipleAlign.append(alignB)
for k, seq in enumerate(multipleAlign):
print(k, seq)
def consensusMultipleAlign(seqs, threshold, simMatrix):
#counts number of sequences
n = len(seqs)
#setting default penalty and new list
insert = 2
extend = 4
multipleAlign = []
i = 0
#looping through sequences
#set j as i+1 to compare first to second seq
for j in range(i + 1, n):
#seqB starts as first sequence
seqB = seqs[j]
#at the beginning, seqA is our first sequence
if not multipleAlign:
seqA = seqs[i]
#Align our two sequences (with gaps!)
#outputs: score and two newly aligned sequences
score, alignA, alignB = sequenceAlign(seqA, seqB, simMatrix,
insert, extend)
#adds newly aligned sequences to our list
multipleAlign.append(alignA)
multipleAlign.append(alignB)
else:
# seqA is our consensus sequence and align seqB to our consensus seq
seqA = consensus(multipleAlign, threshold)
score, alignA, alignB = sequenceAlign(seqA, seqB, simMatrix,
insert, extend)
#creates new list
gaps = []
#for everythin in alignA list, k is index, and letter is value
for k, letter in enumerate(alignA):
# if letter is a dash (there's a gap)
if letter == '-':
#appends index to gap
#keeps track of locations of gaps
gaps.append(k)
#enumerate through each sequence in multipleAlign
for k, seq in enumerate(multipleAlign):
#go through list of gaps
for gap in gaps:
# add a dash at each index for each gap
seq = seq[:gap] + '-' + seq[gap:]
#
multipleAlign[k] = seq
#
multipleAlign.append(alignB)
#
for k, seq in enumerate(multipleAlign):
print(k, seq)
def seqStructureBackboneAlign(chainA,
chainB,
atomNames=set(['CA', 'C', 'N']),
simMatrix=BLOSUM62):
structureA = chainA.structure
structureB = chainB.structure
residuesA = chainA.residues
residuesB = chainB.residues
seqA = getChainSequence(chainA)
seqB = getChainSequence(chainB)
score, alignA, alignB = sequenceAlign(seqA, seqB, simMatrix)
pairedPosA = []
pairedPosB = []
posA = 0
posB = 0
for i in range(len(alignA)):
# No dashes in both at same location
if alignA[i] == '-':
posB += 1
elif alignB[i] == '-':
posA += 1
else:
pairedPosA.append(posA)
pairedPosB.append(posB)
posA += 1
posB += 1
filterIdsA = [residuesA[p].seqId for p in pairedPosA]
filterIdsB = [residuesB[p].seqId for p in pairedPosB]
backboneStrucA = filterSubStructure(structureA, [chainA.code], filterIdsA,
atomNames)
backboneStrucB = filterSubStructure(structureB, [chainB.code], filterIdsB,
atomNames)
atomsA, coordsA = getAtomCoords(backboneStrucA)
atomsB, coordsB = getAtomCoords(backboneStrucB)
weights = ones(len(atomsA))
coordsA, centerA = centerCoords(coordsA, weights)
coordsB, centerB = centerCoords(coordsB, weights)
coords = [coordsA, coordsB]
c, rmsds, atomRmsds, rotations = superimposeCoords(coords, weights)
affineTransformStructure(structureA, rotations[0], -centerA)
affineTransformStructure(structureB, rotations[1], -centerB)
return rmsds, atomRmsds
def seqStructureBackboneAlign(chainA, chainB,
atomNames=set(['CA','C','N']),
simMatrix=BLOSUM62):
structureA = chainA.structure
structureB = chainB.structure
residuesA = chainA.residues
residuesB = chainB.residues
seqA = getChainSequence(chainA)
seqB = getChainSequence(chainB)
score, alignA, alignB = sequenceAlign(seqA, seqB, simMatrix)
pairedPosA = []
pairedPosB = []
posA = 0
posB = 0
for i in range(len(alignA)):
# No dashes in both at same location
if alignA[i] == '-':
posB += 1
elif alignB[i] == '-':
posA += 1
else:
pairedPosA.append(posA)
pairedPosB.append(posB)
posA += 1
posB += 1
filterIdsA = [residuesA[p].seqId for p in pairedPosA]
filterIdsB = [residuesB[p].seqId for p in pairedPosB]
backboneStrucA = filterSubStructure(structureA, [chainA.code],
filterIdsA, atomNames)
backboneStrucB = filterSubStructure(structureB, [chainB.code],
filterIdsB, atomNames)
atomsA, coordsA = getAtomCoords(backboneStrucA)
atomsB, coordsB = getAtomCoords(backboneStrucB)
weights = ones(len(atomsA))
coordsA, centerA = centerCoords(coordsA, weights)
coordsB, centerB = centerCoords(coordsB, weights)
coords = [coordsA, coordsB]
c, rmsds, atomRmsds, rotations = superimposeCoords(coords, weights)
affineTransformStructure(structureA, rotations[0], -centerA)
affineTransformStructure(structureB, rotations[1], -centerB)
return rmsds, atomRmsds
def getDistanceMatrix(seqs, simMatrix):
n = len(seqs)
matrix = [[0.0] * n for x in range(n)]
maxScores = [calcSeqSimilarity(x, x, simMatrix) for x in seqs]
for i in range(n-1):
seqA = seqs[i]
for j in range(i+1,n):
seqB = seqs[j]
score, alignA, alignB = sequenceAlign(seqA, seqB, simMatrix)
maxScore = max(maxScores[i],maxScores[j])
dist = maxScore - score
matrix[i][j] = dist
matrix[j][i] = dist
return matrix
def getDistanceMatrix(seqs, simMatrix):
n = len(seqs)
matrix = [[0.0] * n for x in range(n)]
maxScores = [calcSeqSimilarity(x, x, simMatrix) for x in seqs]
for i in range(n - 1):
seqA = seqs[i]
for j in range(i + 1, n):
seqB = seqs[j]
score, alignA, alignB = sequenceAlign(seqA, seqB, simMatrix)
maxScore = max(maxScores[i], maxScores[j])
dist = maxScore - score
matrix[i][j] = dist
matrix[j][i] = dist
return matrix
def simpleProfileMultipleAlign(seqs, simMatrix):
n = len(seqs)
score, alignA, alignB = sequenceAlign(seqs[0], seqs[1], simMatrix)
multipleAlign = [alignA, alignB]
for i in range(2, n):
profA = profile(multipleAlign)
toAdd = [
seqs[i],
]
profB = profile(toAdd)
score, alignA, alignB = profileAlign(profA, profB, simMatrix)
gaps = []
for j, fractions in enumerate(alignA):
if fractions is None:
gaps.append(j)
for j, seq in enumerate(multipleAlign):
for gap in gaps:
seq = seq[:gap] + '-' + seq[gap:]
multipleAlign[j] = seq
gaps = []
for j, fractions in enumerate(alignB):
if fractions is None:
gaps.append(j)
for j, seq in enumerate(toAdd):
for gap in gaps:
seq = seq[:gap] + '-' + seq[gap:]
toAdd[j] = seq
multipleAlign.extend(toAdd)
return multipleAlign