def suggest_dna(self, best=False, CodonObj=None): """Translates the self.seq protein sequence to DNA If best == False : will pick a random codon with frequency proportional to frequency table If best == True : will always pick the codon with the highest frequency in frequency table """ dna='' if not CodonObj: CodonObj = CodonData() for ltr in self.seq: dna+=CodonObj.codon( ltr, best ) return dna
def __init__( self, frags, start_pos, path ): self.frags = frags self.start_pos = start_pos self.__spos, self.__ph = divmod(start_pos,3) self.DNAobj = None self.CodonData = CodonData() self.min_overlap=8 self.__order_frags() self.assemble( self.min_overlap ) self.__path = path self.__made_mods=[]
class FragLib( object ):
def __init__( self, frags, start_pos, path ):
self.frags = frags
self.start_pos = start_pos
self.__spos, self.__ph = divmod(start_pos,3)
self.DNAobj = None
self.CodonData = CodonData()
self.min_overlap=8
self.__order_frags()
self.assemble( self.min_overlap )
self.__path = path
self.__made_mods=[]
def __order_frags(self):
""" Automatically sorts the fragment list into the proper assembly order
and ensures all the fragments have a logical place
"""
for frag1 in self.frags:
for frag2 in self.frags:
frag1.overlaps(frag2, self.min_overlap)
head=None
tail=None
for frag in self.frags:
if frag.overlap_data[0] == None:
assert( head == None )
head = frag
if frag.overlap_data[1] == None:
assert( tail == None )
tail = frag
ptr=head
tmp_frags=[]
while (ptr != None):
tmp_frags.append(ptr)
ptr = ptr.overlap_data[1]
def algn_file_out(self): #Assumes the fragments were properly shuffled by __order_frags
pos,pos_n,neg, neg_n='','','',''
seq = self.DNAobj.seq
for frag in self.frags:
st_pos = seq.find( frag.seq() )
this_seq = frag.seq(False)
if frag.is_fwd():
diff = st_pos - len(pos)
name_lbl = '|-> '+frag.name + ' +'+str(frag.number)
pos_n += diff*' ' + name_lbl + (' '* ( len(frag)- len(name_lbl) ) )
pos += diff*' ' + this_seq
if frag.is_rev():
diff = st_pos - len(neg)
name_lbl = '|<- ' + frag.name +' -' + str(frag.number)
neg += diff*' ' + this_seq
neg_n += diff*' ' + name_lbl + (' '* ( len(frag)- len(name_lbl) ) )
mut_info=''
for i, one in enumerate(self.__made_mods):
if ( i== 0 ): mut_info += "Mutations: "
mut_info += one+' '
return_data = "vi:set nowrap:\n\n" +'\n'
return_data += self.DNAobj.fancy_print( lw=len(seq)+3 )
return_data += pos_n+'\n'+pos+'\n'+neg+'\n'+neg_n+'\n'
return_data += self.__str__()
h=open( self.__path+'.assembly','w')
h.writelines( return_data )
h.close()
def __str__(self):
ret_str='startpos '+str( 3*self.__spos + self.__ph ) +'\n'
altered=[]
for frag in self.frags:
if frag.updated: altered.append(frag)
ret_str += ( frag.__str__() + '\n' )
if len(altered) != 0:
ret_str += '\n\n'+20*'*'+'NEW FRAGMENTS'+20*'*'+'\n'
for one in altered:
ret_str += one.__str__() +'\n'
return ret_str
def output(self):
fwd_s, rev_s, tot_s = '','', self.DNAobj.seq
data = self.DNAobj.fancy_print( lw = len(tot_s)+ 5 )
altered,al_str=[], ''
for frag in self.frags:
align = tot_s.find( frag.seq() )
if frag.is_fwd():
pad = ' '*( align - len(fwd_s) )
fwd_s += ( pad + frag.seq() )
else:
pad = ' '*( align - len(rev_s) )
rev_s += ( pad + frag._Fragment__seq )
if frag.updated: altered.append(frag)
return 'vi:set nowrap:'+'\n'+ data+fwd_s+'\n'+rev_s+'\n\n\n'+self.__str__()
def assemble( self, min_overlap = 6 ):
seq=''
offset=0
for i,frag in enumerate(self.frags):
frag.coding_codon={}
if i == 0:
seq += frag.seq().replace(' ','')
codons = split_codons( frag.seq(), self.__ph )
##find where to start counting codons:
offset = self.__spos
if ( self.__ph != 0 ): offset += 1
##End counting logic block
for indx, codon in enumerate(codons):
frag.coding_codon[ 1 + indx - offset ]=codon
continue
overlap = seq.find( frag.seq()[0:min_overlap] )
assert( overlap != -1 )
if (seq[overlap:] != frag.seq()[0:(len(seq)-overlap)] ): ##CHECK THAT WE HAVE PERFECT ALIGNMENT!
#get required info
f1obj = self.frags[i-1]
f2obj = frag
f1name = f1obj.name + f1obj._Fragment__polarity +f1obj.number
f2name = f2obj.name + f2obj._Fragment__polarity +f2obj.number
print "Mismatch between framgnets "
print ljust(f1name,20), f1obj.seq(False)
print ljust(f2name,20), f2obj.seq(False)
sys.exit()
seq += frag.seq()[(len(seq) - overlap):]
num,ph = divmod( overlap-self.start_pos, 3)
ph=3-ph
codons = split_codons( frag.seq(), ph )
for ii, codon in enumerate(codons):
this_index = num + ii + 1
frag.coding_codon[ this_index ] = codon
##########Optional block, double check for consistancy#######
if self.frags[i-1].contains(this_index):
other = self.frags[i-1].coding_codon[this_index]
if (' ' in other) or (' ' in codon): continue
if other != codon: print this_index, other, codon
self.DNAobj = DNASeq( seq, self.start_pos )
def mutate(self, key):
AA1,num,AA2 = key.split('.')
if AA2 != '': the_codon = self.CodonData.codon(AA2, False)
else: the_codon = '' #Covers deletions
try: self.DNAobj.fetch_position( int(num)-1, AA1 ) ##Sanity check, ensure AA1 is indeed at pos num
except WrongAA as e:
sys.exit("Resiude number "+str(num)+" is "+e.AA+" not an "+AA1+"! Please check mutation input")
for frag in self.frags:
if not frag.contains( num ): continue
frag.mutate(int(num),the_codon)
if AA2 !='': frag.name+= '_'+AA1+str(num)+AA2
else: frag.name+= '_x'+AA1+str(num)
self.assemble( self.min_overlap )
self.__made_mods.append(key)
def insert(self, pos, AAseq):
new_dna_seq, new_frag_seq = '',''
for aa in AAseq: new_dna_seq+= self.CodonData.codon(aa, False)
for frag in self.frags:
if not frag.contains(pos): continue
if ' ' == frag.coding_codon[int(pos)][:-1]: continue
for cod in sorted(frag.coding_codon.keys()):
new_frag_seq += frag.coding_codon[cod]
if cod == int(pos): new_frag_seq += new_dna_seq
if frag.is_rev(): new_frag_seq = complement(new_frag_seq.replace(' ','') )
frag._Fragment__seq = new_frag_seq
self.assemble(self.min_overlap )
