def runMafftAlignmentWithMatrix(inputSequences,
gap_open,
gap_ext,
method="globalpair",
matrixfile=""):
""" Run MAFFT alignment on a group of sequences with a substitution matrix.
This function requires a proper path to the mafft binary.
Args:
inputSequences (array): an array of strings of sequences to be aligned
gap_open (float): gap open penalty
gap_ext (float): gap extent penalty
method (str): MAFFT's alignment strategy
matrixfile (str): path to an appropriate matrix file for MAFFT
Returns:
output (array): aligned strings of sequences
"""
tempfile = "temp/tempseq.fasta"
tempfileout = "temp/tempmsa.fasta"
tempfileout2 = "temp/tempmsa2.fasta"
io.exportGroupOfSequencesToFASTA(inputSequences, tempfile)
command = "mafft-mac/mafft.bat --quiet --op " + str(
gap_open
) + " --ep " + str(
gap_ext
) + " --" + method + " --maxiterate 1000 --aamatrix " + matrixfile + " " + tempfile + " > " + tempfileout2
os.system(command)
output = io.readFASTA(tempfileout2)
return output
def runMafftAlignmentWithSettings(inputSequences,
gap_open,
gap_ext,
method="globalpair",
allowshift=False):
""" Run MAFFT alignment on a group of sequences without a substitution matrix.
This function requires a proper path to the mafft binary.
Args:
inputSequences (array): an array of strings of sequences to be aligned
gap_open (float): gap open penalty
gap_ext (float): gap extent penalty
method (str): MAFFT's alignment strategy
allowshift (bool): see MAFFT documentation
Returns:
output (array): aligned strings of sequences
"""
tempfile = "temp/tempseq.fasta"
tempfileout = "temp/tempmsa.fasta"
tempfileout2 = "temp/tempmsa2.fasta"
io.exportGroupOfSequencesToFASTA(inputSequences, tempfile)
if allowshift == True:
command = "mafft-mac/mafft.bat --quiet --op " + str(
gap_open
) + " --ep " + str(
gap_ext
) + " --" + method + " --maxiterate 1000 --allowshift --text " + tempfile + " > " + tempfileout2
else:
command = "mafft-mac/mafft.bat --quiet --op " + str(
gap_open
) + " --ep " + str(
gap_ext
) + " --" + method + " --maxiterate 1000 --text " + tempfile + " > " + tempfileout2
os.system(command)
output = io.readFASTA(tempfileout2)
return output
def tcoffeeAlignment(sequences, go, ge):
""" The standard T-COFFEE alignment.
For more info: http://www.tcoffee.org/Projects/tcoffee/#DOCUMENTATION
Args:
sequences (array): array of strings to be aligned
go (int): gap open penalty
ge (int): gap extent penalty
Returns:
output (array): aligned sequences
"""
io.exportGroupOfSequencesToFasta(sequences, "temp/tcoffee-in.fasta")
command = "./t_coffee 'temp/tcoffee-in.fasta' -quiet -output fasta -outfile 'temp/tcoffee-out.fasta'"
os.system(command)
output = io.readFASTA("temp/tcoffee-out.fasta")
return output
def tcoffeeAlignmentWithMatrix(sequences, go, ge, matrixfile):
""" The standard T-COFFEE alignment with custom matrix
For more info: http://www.tcoffee.org/Projects/tcoffee/#DOCUMENTATION
Args:
sequences (array): array of strings to be aligned
go (int): gap open penalty
ge (int): gap extent penalty
matrixfile (str): file path to an appropriate matrix file
Returns:
output (array): aligned sequences
"""
io.exportGroupOfSequencesToFasta(sequences, "temp/tcoffee-in.fasta")
command = "./t_coffee 'temp/tcoffee-in.fasta' -in = X" + matrixfile + " -gapopen = " + str(
go) + " -gapext = " + str(
ge) + " -quiet -output fasta -outfile 'temp/tcoffee-out.fasta'"
os.system(command)
output = io.readFASTA("temp/tcoffee-out.fasta")
return output
# align them using MAFFT
MSA = al.runMafftAlignmentWithSettings(sequences,
2,
1,
method="globalpair",
allowshift=False)
print "Aligned sequences:"
al.printMSA(MSA)
# Create a Krogh profile HMM from a folk tune family
# First we need the emission probabilities gathered from the whole dataset
directory = "NotAligned/NLBproperSmall"
files = io.filesInPath(directory)
allPossibleSymbols = "".join([
"".join(["".join(seq) for seq in io.readFASTA(directory + "/" + file)])
for file in files
]).replace("-", "")
counts = collections.Counter(allPossibleSymbols) # count
alphabet = [symbol for symbol in counts] # all possible symbols
counts = np.array([counts[symbol]
for symbol in counts]) # corresponding counts
EmissionProbabilities = dict(zip(alphabet, counts /
np.sum(counts))) # convert to dictionary
# Bake the Krogh profile HMM model
sw = 10
pct = 1
pec = 1 / 12.
grms = 0.95
model, columnsIsMatchState = krogh.profileHMM(