def _buildMatchVersions(self, miRNA_seq):
"""
"""
returnDict = {}
# Extract all versions of seed from miRNA
for seedType in _seedModels:
returnDict[seedType] = []
# If NO third index like here: [0,7,'A1']
if len(_seedModels[seedType]) == 2:
index_0 = _seedModels[seedType][0]
index_1 = _seedModels[seedType][1]
returnDict[seedType].append(bioDefs.revComp(miRNA_seq[index_0:index_1].replace('U','T')))
# If extra data: use 'instructions index' to construct seedVersion
elif len(_seedModels[seedType]) == 3:
nuc,pos = list(_seedModels[seedType][2])
pos = int(pos)
index_0 = _seedModels[seedType][0]
index_1 = _seedModels[seedType][1]
returnDict[seedType].append(bioDefs.revComp(miRNA_seq[index_0:index_1].replace('U','T')))
# Convert correct position to reflect adjustment listed
# in 'instructions'.
returnDict[seedType][0] = list(returnDict[seedType][0]) # convert to list for inplace mutation
returnDict[seedType][0][-pos] = nuc # insert nuc
returnDict[seedType][0] = ''.join(returnDict[seedType][0]) # convert back to string
return returnDict
def _loadSeed(self):
# Determine whether we got miRNA or kmer and initialize
# 'seed' and 'seedComp' dicts.
if len(self.sourceSeq) > 8:
self.sourceSeqType = 'miRNA'
# Extract all versions of seed from miRNA
for seedType in self.seedModels:
# If no extra info take the defined slice
if len(self.seedModels[seedType]) == 2:
index_0 = self.seedModels[seedType][0]
index_1 = self.seedModels[seedType][1]
self.seed[seedType] = self.sourceSeq[index_0:index_1].replace('U','T')
# If extra data: use 'instructions' to construct seed where
# revComp will generate match to message.
elif len(self.seedModels[seedType]) == 3:
nuc,pos = list(self.seedModels[seedType][2])
pos = int(pos)
index_0 = self.seedModels[seedType][0]
index_1 = self.seedModels[seedType][1]
self.seed[seedType] = self.sourceSeq[index_0:index_1].replace('U','T')
# Convert correct position to reflect adjustment listed
# in 'instructions'.
self.seed[seedType] = list(self.seed[seedType]) # convert to list for inplace mutation
self.seed[seedType][pos-1] = bioDefs.revComp(nuc) # insert rvCmp'd nuc
self.seed[seedType] = ''.join(self.seed[seedType]) # convert back to string
else:
self.sourceSeqType = 'kmer'
self.seed['kmer'] = self.sourceSeq.replace('U','T')
def _loadmatchVersions(self, set4realMatches):
# Determine whether we got miRNA or kmer and initialize
# matchVersions Dict.
assert self.sourceSeq > 8, \
'ERROR: self.sourceSeq (%s) must be > 8 nt long.' % (self.sourceSeq)
self.sourceSeqType = 'miRNA'
# We use the m2_to_m8 version to represent the seedVersion set for an miRNA since all versions can be constructed from this one.
set4realMatches.add(bioDefs.revComp(self.sourceSeq[1:8].replace('U','T')))
matchVersions = self._buildMatchVersions(self.sourceSeq)
for seedType in _seedModels:
self.matchVersions[seedType] = matchVersions[seedType]
# Log each version of the seed match to use as restrictedList when building Ctrls
set4realMatches.add(matchVersions[seedType][0]) # for technical reasons matchVersions[seedType] == ['aSeedMatchSeq'] so we must index it
def buildCtrlsFromProSeed(self, restrictedList, numOfCtrls=15):
"""
WARNING: This should only be called after the entire list of real seeds have been initialized!
Computes 15 permutations of the 'true' proSeed matching seqeunce (m2_to_m8) and derives
matchVersions as in the true case. The permuted sequence is checked agaisnt the restrictedList
to prevent using KNOWN seed matches for controls. Ctrl seed matches are stored in a list located
at second index of the list located in dict entry self.matchVersions[seedType]. Each seedVersion
of a ctrl set will share the same index number.
"""
##assert True==False, \
##"""WARNING!!! miRNA.buildCtrlsFromMatchVers() should be used instead!!
##If you REALLY want to use this method, modify or remove this assert statement.
##But seriously... use miRNA.buildCtrlsFromMatchVers()."""
# check to see whether this has already been done.
# If so, complain and die.
# Else, append an empty list as index_1 after REAL matchSeq for each version
for seedType in self.matchVersions:
assert type(self.matchVersions[seedType]) == type([]), \
'ERROR: %s.matchVersions[%s] is not type: list.' % (self.name,seedType)
assert len(self.matchVersions[seedType]) == 1, \
'ERROR: len(%s.matchVersions[%s]) is not 1; ctrls seqs may have already been built.' % (self.name,seedType)
self.matchVersions[seedType].append([])
proSeed = self.sourceSeq[1:8]
matchPerms = [''.join(x) for x in xpermutations.xpermutations(list(self.matchVersions['m2_to_m8'][0]))]
# Select 15 random permutations of matchVersions['m2_to_m8'] that are not in the
# restrictedList.
chosenPerms = []
while len(chosenPerms) < numOfCtrls:
permSeq = JamesDefs.randFromList_noReplace(matchPerms)
if permSeq not in restrictedList:
chosenPerms.append(permSeq)
# Use each chosenSeq to generate the diff matchVersions
for seq in chosenPerms:
# Create Fake miRNA with seq at the seed location to feed to _buildMatchVersions()
seq = 'N%sNNNNNNNNNNNNN' % (bioDefs.revComp(seq))
matchVersions = self._buildMatchVersions(seq)
for seedType in self.matchVersions:
self.matchVersions[seedType][1].append(matchVersions[seedType][0]) # must use index[0] bc _buildMatchVersions returns a list len==1
oligoType = 'control' # 'match' or 'control'
assert oligoType == 'match' or 'control', 'oligoType MUST be only "match" or "control".'
# Load miRNA fastas into dict.
miRNAs = Fasta.file2dict(miRNAFile)
# Create new dict for seeds.
seeds = {}
# 1) Cycle through miRNA dict taking 7mers starting at pos 1
# and then pos2. Adapt key to reflect which.
# 2) Convert to all uppers and convert U's to T's
# 3) If oligoType == 'match', rvcmp each 7mer and adapt key
# to reflect which.
for miRNA in miRNAs:
pos1_seed = miRNAs[miRNA][:7].upper().replace('U','T')
pos2_seed = miRNAs[miRNA][1:8].upper().replace('U','T')
if oligoType == 'match':
seeds[miRNA+'_match_pos1'] = bioDefs.revComp(pos1_seed)
seeds[miRNA+'_match_pos2'] = bioDefs.revComp(pos2_seed)
else:
seeds[miRNA+'_ctrl_pos1'] = pos1_seed
seeds[miRNA+'_ctrl_pos2'] = pos2_seed
# Write out seed dict as fasta.
Fasta.write(seeds,seedFile)
print "Done."
from gusPyCode.defs.bioDefs import revComp
# To convert Aa putative TSS seqs into +1 Ori based on gene's Ori
chromOriSeqs = map(lambda line: line.strip(), open('/Users/biggus/Documents/James/Data/AedesCorePromoterWork/output/Inr_DPE/Aa_Ensbl49_AaegL1.1.plus100minus100._InrDPE_.newInr.chromStrand.fa','rU').readlines())
outFile = '/Users/biggus/Documents/James/Data/AedesCorePromoterWork/output/Inr_DPE/Aa_Ensbl49_AaegL1.1.plus100minus100._InrDPE_.newInr.geneStrand.local.fa'
geneOriSeqs = []
chromOriSeqs = zipped = zip(chromOriSeqs[:-1:2], chromOriSeqs[1::2])
for item in chromOriSeqs:
fieldsList = item[0].split(' ')
seq = item[1]
if fieldsList[-1] == '(1:-1)':
fieldsList[-1] = '(-1:1)'
seq = revComp(seq)
geneOriSeqs.append(' '.join(fieldsList)+'\n')
geneOriSeqs.append(seq+'\n')
else:
geneOriSeqs.append(item[0]+'\n')
geneOriSeqs.append(item[1]+'\n')
outFile = open(outFile, 'w')
outFile.writelines(geneOriSeqs)
print 'Done.'
def _loadSeedComp(self):
for seedType in self.seed:
self.seedComp[seedType] = bioDefs.revComp(self.seed[seedType])
from TAMO.seq import Fasta
from gusPyCode.defs.bioDefs import revComp
seqFile = '/Users/biggus/Documents/James/Collaborations/Campbell/data/mainTwoGenes.fas'
kmerFile = '/Users/biggus/Documents/James/Collaborations/Campbell/data/mainTwoGenes.7mersInAll.tamoVers.txt'
#outFile = '/Users/biggus/Documents/James/Collaborations/Campbell/data/mainTwoGenes.7mersInAll.tamoVers.txt'
seqs = Fasta.file2dict(seqFile)
seqNames = sorted(seqs.keys())
kmers = map(lambda l: l.strip(), open(kmerFile, 'rU').readlines())
kmers.append('AAHRRSSSSSSSSSMMMMM')
results = []
for kmer in kmers:
print '>%s:' % (kmer)
results.append('%s:' % (kmer))
for seq in seqNames:
if seqs[seq].find(kmer) != -1 or seqs[seq].find(revComp(kmer)) != -1:
print '\t'+seq
results.append(seq)
else:
print '---NOT FOUND---'
#outFile = open(outFile, 'w')
#for each in inAllSeqs:
#outFile.write(each+'\n')
print 'Done'
