def run_meme(seqs, outdir, meme_settings):
"""Motif discovery on given sequences via MEME. Return MEME record"""
infile = tempfile.NamedTemporaryFile()
with open(infile.name, 'w') as f:
f.write(seqs_to_fasta(seqs))
# Call shell command
cmd = [
meme_settings['path'],
infile.name,
'-oc',
outdir,
'-dna', # sequences use DNA alphabet
'-revcomp', # allow sites on both strands
'-maxsize',
10000000
] # max dataset size on chars
# extend meme command
for opt in ['cons', 'nmotifs', 'mod', 'minw', 'maxw']:
if opt in meme_settings:
cmd.extend(['-' + opt, meme_settings[opt]])
cmd = map(str, cmd)
#print "Running MEME."
print ' '.join(cmd)
call(cmd)
# Parse text output of the MEME
with open(os.path.join(outdir, 'meme.txt')) as f:
rec = MEME.read(f)
return rec
def import_from_MEME(self,filename,n=1,mode='biotools'):
"""imports a motif from the output of MEME (meme.txt)
if there are multiple motifs in the output, we will use motif n (the first is n=1, which is also the default)
"""
import Bio.Motif.Parsers.MEME as MEME
f = open(filename)
MEME_object = MEME.read(f)
motif_name = 'Motif ' + str(n)
biopython_motif = MEME_object.get_motif_by_name(motif_name)
if mode=='biopython': return biopython_motif
if mode=='biotools':
internal_n = len(biopython_motif)
# this next line is instead of initializePositions
biotools_motif = [CharDict(biopython_motif[i]) for i in range(internal_n)]
self._L = biotools_motif
self._n = internal_n
else: raise UserWarning('Not a valid mode.')
def parse(handle, format):
"""Parses an output file of motif finding programs.
Currently supported formats:
- AlignAce: AlignAce output file format
- MEME: MEME output file motif
- TRANSFAC: TRANSFAC database file format
- pfm: JASPAR-style position-frequency matrix
- sites: JASPAR-style sites file
- jaspar-pfm: JASPAR-style position-frequency matrix [DEPRECATED]
- jaspar-sites: JASPAR-style sites file [DEPRECATED]
As files in the pfm and sites formats contain only a single motif,
it is easier to use Bio.Motif.read() instead of Bio.Motif.parse()
for those.
For example:
>>> from Bio import Motif
>>> for motif in Motif.parse(open("Motif/alignace.out"),"AlignAce"):
... print motif.consensus()
TCTACGATTGAG
CTGCACCTAGCTACGAGTGAG
GTGCCCTAAGCATACTAGGCG
GCCACTAGCAGAGCAGGGGGC
CGACTCAGAGGTT
CCACGCTAAGAGAAGTGCCGGAG
GCACGTCCCTGAGCA
GTCCATCGCAAAGCGTGGGGC
GAGATCAGAGGGCCG
TGGACGCGGGG
GACCAGAGCCTCGCATGGGGG
AGCGCGCGTG
GCCGGTTGCTGTTCATTAGG
ACCGACGGCAGCTAAAAGGG
GACGCCGGGGAT
CGACTCGCGCTTACAAGG
>>> for motif in Motif.parse(open("Motif/alignace.out"),"alignace"):
... print motif.consensus
TCTACGATTGAG
CTGCAGCTAGCTACGAGTGAG
GTGCTCTAAGCATAGTAGGCG
GCCACTAGCAGAGCAGGGGGC
CGACTCAGAGGTT
CCACGCTAAGAGAGGTGCCGGAG
GCGCGTCGCTGAGCA
GTCCATCGCAAAGCGTGGGGC
GGGATCAGAGGGCCG
TGGAGGCGGGG
GACCAGAGCTTCGCATGGGGG
GGCGTGCGTG
GCTGGTTGCTGTTCATTAGG
GCCGGCGGCAGCTAAAAGGG
GAGGCCGGGGAT
CGACTCGTGCTTAGAAGG
"""
if format=="AlignAce":
# Old Motif code
from Bio.Motif.Parsers import AlignAce
record = AlignAce.read(handle)
return iter(record.motifs)
elif format=="alignace":
# Old Motif code
from Bio.Motif import AlignAce
record = AlignAce.read(handle)
return record
elif format=="MEME":
from Bio.Motif.Parsers import MEME
record = MEME.read(handle)
return iter(record.motifs)
elif format=="meme":
from Bio.Motif import MEME
record = MEME.read(handle)
return record
elif format=="TRANSFAC":
from Bio.Motif import TRANSFAC
record = TRANSFAC.read(handle)
return record
elif format in ('pfm', 'sites'):
from Bio.Motif import Jaspar
motif = Jaspar.read(handle, format)
elif format=="jaspar-pfm":
motif = OldMotif()._from_jaspar_pfm(handle)
return iter([motif])
elif format=="jaspar-sites":
motif = OldMotif()._from_jaspar_sites(handle)
return iter([motif])
else:
raise ValueError("Unknown format %s" % format)
# Treat the single-motif formats
motifs = [motif]
return motifs
fin = open(fasta_sequences, "r")
fout = open(fasta_with_weights, "w")
for line in fin:
fout.write( line.replace (">WEIGHTS", weights_string) )
fin.close()
fout.close()
#call command line meme
subprocess.call(["meme", "-dna", "-revcomp", "seqweights.fasta", "-minw", "8", "-maxw", "14"])
os.chdir(output_directory)
f = open("meme.txt")
record = MEME.read(f)
for motif in record.motifs:
for instance in motif.instances:
if instance.sequence_name == myseq:
print instance.sequence_name, instance.start, instance.length
found += 1
for i in range(instance.length):
position = instance.start+i - 1 #subtract one to adjust for meme indexing from 1 rather than 0
promoter_array[position] += 1
else:
print "not found"
not_found += 1
os.chdir(current_directory)
percent_found = found / (found+not_found)
