0
|
1 #This module collects functions used in homology comparison and frame prediction
|
|
2
|
|
3 import subprocess
|
|
4
|
|
5 def maptranscript(gene,exon_end,part=''):
|
|
6 """
|
|
7 Find all transcripts of the gene having the fusion boundary.
|
|
8 "gene" is a Gene object.
|
|
9 "exon_end" is the exon boundary that to be searched.
|
|
10 "part" has to be '5' or '3'.
|
|
11 Return a list of Transcript objects that has "exon_end" boundary.
|
|
12 """
|
|
13 if part not in ['5','3']:
|
|
14 raise Exception('part must be "5" or "3"')
|
|
15 txuse=[] ##
|
|
16 g=gene.name
|
|
17 strand=gene.strand
|
|
18 if part=='5':
|
|
19 tag='big' if strand=='1' else 'small'
|
|
20 else:
|
|
21 tag='small' if strand=='1' else 'big'
|
|
22 txs=gene.transcript
|
|
23 if tag=='small':
|
|
24 for tx in txs:
|
|
25 e_pos=[x.start for x in tx.exon]
|
|
26 if exon_end in e_pos:
|
|
27 txuse.append(tx)
|
|
28 elif tag=='big':
|
|
29 for tx in txs:
|
|
30 e_pos=[x.end for x in tx.exon]
|
|
31 if exon_end in e_pos:
|
|
32 txuse.append(tx)
|
|
33 return txuse
|
|
34
|
|
35
|
|
36 def overlap(r1,r2):
|
|
37 """compare two ranges, return the overlapped range"""
|
|
38 if r1[0] <= r2[0]:
|
|
39 ra,rb=r1[:],r2[:]
|
|
40 else:
|
|
41 ra,rb=r2[:],r1[:]
|
|
42 if rb[0] > ra[1]:
|
|
43 return []
|
|
44 else:
|
|
45 if rb[1] <= ra[1]:
|
|
46 return rb
|
|
47 else:
|
|
48 return [rb[0],ra[1]]
|
|
49
|
|
50 def br_front_len(tx,br,part):
|
|
51 '''No matter it is 5'/3' partner,calculate the ORF length before the break'''
|
|
52 txname=tx.name
|
|
53 strand=tx.strand
|
|
54 cds_start=tx.cds_start
|
|
55 cds_end=tx.cds_end
|
|
56 e_pos=[[x.start,x.end] for x in tx.exon]
|
|
57 if part=='5' and strand=='1':
|
|
58 posset=[x.end for x in tx.exon]
|
|
59 if br not in posset:
|
|
60 raise Exception('Br is not exon boundary')
|
|
61 if br < cds_start:
|
|
62 return '5UTR'
|
|
63 elif br > cds_end:
|
|
64 return '3UTR'
|
|
65 else:
|
|
66 chim_r=[cds_start,br]
|
|
67 ol=[overlap(x,chim_r) for x in e_pos]
|
|
68 L=sum([x[1]-x[0]+1 for x in ol if len(x)>0])
|
|
69 return L
|
|
70 if part=='5' and strand=='-1':
|
|
71 posset=[x.start for x in tx.exon]
|
|
72 if br not in posset:
|
|
73 raise Exception('Br is not exon boundary')
|
|
74 if br > cds_end:
|
|
75 return '5UTR'
|
|
76 elif br < cds_start:
|
|
77 return '3UTR'
|
|
78 else:
|
|
79 chim_r=[br,cds_end]
|
|
80 ol=[overlap(x,chim_r) for x in e_pos]
|
|
81 L=sum([x[1]-x[0]+1 for x in ol if len(x)>0])
|
|
82 return L
|
|
83 if part=='3' and strand=='1':
|
|
84 posset=[x.start for x in tx.exon]
|
|
85 if br not in posset:
|
|
86 raise Exception('Br is not exon boundary')
|
|
87 if br < cds_start:
|
|
88 return '5UTR'
|
|
89 elif br > cds_end:
|
|
90 return '3UTR'
|
|
91 else:
|
|
92 chim_r=[cds_start,br-1]
|
|
93 ol=[overlap(x,chim_r) for x in e_pos]
|
|
94 L=sum([x[1]-x[0]+1 for x in ol if len(x)>0])
|
|
95 return L
|
|
96 if part=='3' and strand=='-1':
|
|
97 posset=[x.end for x in tx.exon]
|
|
98 if br not in posset:
|
|
99 raise Exception('Br is not exon boundary')
|
|
100 if br > cds_end:
|
|
101 return '5UTR'
|
|
102 elif br < cds_start:
|
|
103 return '3UTR'
|
|
104 else:
|
|
105 chim_r=[br+1,cds_end]
|
|
106 ol=[overlap(x,chim_r) for x in e_pos]
|
|
107 L=sum([x[1]-x[0]+1 for x in ol if len(x)>0])
|
|
108 return L
|
|
109
|
|
110 def fusion_frame(gene_a,ga_br,gene_b,gb_br):
|
|
111 """gene_a:FGFR3,gene_b:TACC3,ga_br:1808661,gb_br:1741429
|
|
112 or chr4.1808661.chr4.1741429
|
|
113 gene_a, gene_b are Gene objects.
|
|
114 """
|
|
115 ga_br=int(ga_br)
|
|
116 gb_br=int(gb_br)
|
|
117 ga_txs=maptranscript(gene_a,ga_br,part='5')
|
|
118 gb_txs=maptranscript(gene_b,gb_br,part='3')
|
|
119 res=[]
|
|
120 for ta in ga_txs:
|
|
121 for tb in gb_txs:
|
|
122 s1=br_front_len(ta,ga_br,'5')
|
|
123 s2=br_front_len(tb,gb_br,'3')
|
|
124 fusion_conseq='Unknown'
|
|
125 if isinstance(s1,int) and isinstance(s2,int): #both br in CDS region
|
|
126 if s1%3==s2%3:
|
|
127 fusion_conseq='In-frame'
|
|
128 else:
|
|
129 fusion_conseq='Out-of-frame'
|
|
130 elif isinstance(s1,int) and not isinstance(s2,int):
|
|
131 fusion_conseq='CDS'+'-'+s2
|
|
132 elif isinstance(s2,int) and not isinstance(s1,int):
|
|
133 fusion_conseq=s1+'-'+'CDS'
|
|
134 else:
|
|
135 fusion_conseq=s1+'-'+s2
|
|
136 res.append((ta.name,tb.name,fusion_conseq))
|
|
137 if ga_txs==[] and gb_txs==[]:
|
|
138 res.append(['NA','NA','NA'])
|
|
139 elif ga_txs==[]:
|
|
140 for tb in gb_txs:
|
|
141 res.append(['NA',tb.name,'NA'])
|
|
142 elif gb_txs==[]:
|
|
143 for ta in ga_txs:
|
|
144 res.append([ta.name,'NA','NA'])
|
|
145 return res
|
|
146
|
|
147
|
|
148 def seqblast(seqA,seqB,blastn=None):
|
|
149 '''seqA,seqB:fasta files'''
|
|
150 if blastn==None: blastn='blastn'
|
|
151 cmdstr='%s -task=blastn -subject %s -query %s -outfmt 6'%(blastn,seqA,seqB)
|
|
152 cmdout=subprocess.Popen(cmdstr.split(),stdout=subprocess.PIPE,stderr=subprocess.STDOUT)
|
|
153 result=cmdout.stdout.read()
|
|
154 best_align=result.split('\n')[0]
|
|
155 if best_align=='':
|
|
156 return None
|
|
157 else:
|
|
158 info=best_align.split('\t')
|
|
159 identity=info[2].strip()
|
|
160 align_len=info[3].strip()
|
|
161 evalue=info[10].strip()
|
|
162 bit_score=info[11].strip()
|
|
163 return [identity,align_len,evalue,bit_score]
|
|
164
|
|
165
|