def parsing_each_query(content,fout):
#1. have to be 100 similarity
#2. different segment add up together to be the complete length
#3. check strand constrains
#4. check no overlapping constrains
content = string.strip(string.join(content,''))
if content =='':
return
parse_dic = blast_parse.parse_query_alignment(content)
query = parse_dic.keys()[0]
print query,
hit_candidate_segment =[] #[[hitid, strand, pstart, pend, hstart, hend],...]
group_list =[]
query_len= parse_dic[query][0]
hit_dic = parse_dic[query][1] #only two
hit_list = hit_dic.keys()
for hitid in hit_list:
#to get the id
hit_align_list = hit_dic[hitid]
first_100 =0
for seg_list in hit_align_list:
strand, qstart, qend, hstart, hend, percentage = seg_list
if percentage < 100 :#99.5:
continue
if abs(qend - qstart) +1 <query_len:
continue
hit_candidate_segment.append([hitid, strand,qstart,qend, hstart,hend])
#write to file of the group list
fout.write('['+query+']\n')
#write the query itself
fout.write(query+'\n')
fout.write("%s %s %s %s %s\n" % ( '+','1',str(query_len),'1',str(query_len)))
hit_dic={}
if gfdir !="":
string_start = len(gfdir)+1
else:
string_start = 0
for entry in hit_candidate_segment:
hitid, hitstrand, qstart,qend, hstart,hend = entry
hitid = hitid[string_start:]
if gfdir !="":
string_end = string.find(hitid, ".nib:")
hitid = hitid[:string_end]
fout.write(hitid+'\n')
fout.write("%s %s %s %s %s\n" % (hitstrand, str(qstart), str(qend), str(hstart), str(hend)))
fout.write('\n')
print "done"
def parsing_each_query(content,fout):
#1. have to be 100 similarity
#2. different segment add up together to be the complete length
#3. check strand constrains
#4. check no overlapping constrains
content = string.strip(string.join(content,''))
if content =='':
return
parse_dic = blast_parse.parse_query_alignment(content)
query = parse_dic.keys()[0]
print query,
hit_candidate_segment =[] #[[hitid, strand, pstart, pend, hstart, hend],...]
group_list =[]
query_len= parse_dic[query][0]
hit_dic = parse_dic[query][1] #only two
hit_list = hit_dic.keys()
for hitid in hit_list:
#to get the id
hit_align_list = hit_dic[hitid]
first_100 =0
for seg_list in hit_align_list:
strand, qstart, qend, hstart, hend, percentage = seg_list
if percentage < 100 :#99.5:
continue
if first_100 == 0:
first_100 =1
#inital test to see if the hits is likely to be real - large first hsp 100%
if abs(qend - qstart) <50:
break
hit_candidate_segment.append([hitid, strand,qstart,qend, hstart,hend])
parse_dic ={} #save memory only one enty in the parse_dic
#generate the combination constrains dictionary using the list position as the segment
#id to form the dic as id:okid_list
rule_dic, must_cover_list = rules(hit_candidate_segment,query_len)
#generate all the combinations that do not conflict with the rule_dic
combo_list = generate_combination(rule_dic, must_cover_list, hit_candidate_segment, query_len)
#now check if any combination of the segment make a complete query
# also need to check strand, overlapping constrains.
#[[hitid, strand, pstart, pend, hstart, hend],...]
#use 1,2,3,... segment of the list to reconstruct and then check constrains
for i in range(0, len(combo_list)):
candidate_list =[]
for pos in combo_list[i]:
candidate_list.append( hit_candidate_segment[pos])
passConstrain = constrains(query_len,candidate_list)
if passConstrain:
group_list.extend(candidate_list)
#write to file of the group list
fout.write('['+query+']\n')
#write the query itself
fout.write(query+'\n')
fout.write("%s %s %s %s %s\n" % ( '+','1',str(query_len),'1',str(query_len)))
hit_dic={}
if gfdir !="":
string_start = len(gfdir)+1
else:
string_start = 0
for entry in group_list:
hitid, hitstrand, qstart,qend, hstart,hend = entry
hitid = hitid[string_start:]
if gfdir !="":
string_end = string.find(hitid, ".nib:")
hitid = hitid[:string_end]
fout.write(hitid+'\n')
fout.write("%s %s %s %s %s\n" % (hitstrand, str(qstart), str(qend), str(hstart), str(hend)))
fout.write('\n')
print "done"