Mercurial > repos > drosofff > blast_to_scaffold

comparison blast_to_scaffold.py @ 2:61731b911f4e draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload for repository https://github.com/ARTbio/tools-artbio/tree/master/tools/blast_to_scaffold commit e67f150062f2076d0879bcdd3376ef53ba639d91
author drosofff
date Fri, 25 Mar 2016 17:01:52 -0400
parents 7fb07a6ab968
children e279e53a3fde
comparison
equal deleted inserted replaced
1:35e2e6452cf7 2:61731b911f4e
28 '--guideSequence', action="store", type=str, help="the reference sequence to guide the scaffold assembly in fasta format") 28 '--guideSequence', action="store", type=str, help="the reference sequence to guide the scaffold assembly in fasta format")
29 the_parser.add_argument( 29 the_parser.add_argument(
30 '--blast-tab', dest="blast_tab", action="store", type=str, help="13-columns tabular blastn or tblastx output") 30 '--blast-tab', dest="blast_tab", action="store", type=str, help="13-columns tabular blastn or tblastx output")
31 the_parser.add_argument( 31 the_parser.add_argument(
32 '--output', action="store", type=str, help="output file path, fasta format") 32 '--output', action="store", type=str, help="output file path, fasta format")
33 the_parser.add_argument(
34 '--scaffold_prefix', action="store", type=str, help="the prefix that will be used for the header of the fasta scaffold")
35 the_parser.add_argument(
36 '--scaffold_suffix', action="store", type=str, help="the sufix that will be used for the header of the fasta scaffold")
33 args = the_parser.parse_args() 37 args = the_parser.parse_args()
34 return args 38 return args
35 39
36 def blatnInfo (file): 40 def blatnInfo (file):
37 blastlist = [] 41 blastlist = []
96 except KeyError: 100 except KeyError:
97 continue 101 continue
98 for i, nucleotide in enumerate(sequence): 102 for i, nucleotide in enumerate(sequence):
99 GuideDict[i+subjectStart] = nucleotide 103 GuideDict[i+subjectStart] = nucleotide
100 104
101 def finalAssembly (GuideDict, outputfile): 105 def finalAssembly (GuideDict, outputfile, prefix, suffix):
102 finalSeqList = [] 106 finalSeqList = []
103 for keys in sorted(GuideDict): 107 for keys in sorted(GuideDict):
104 finalSeqList.append(GuideDict[keys]) 108 finalSeqList.append(GuideDict[keys])
105 finalSequence = insert_newlines("".join(finalSeqList) ) 109 finalSequence = insert_newlines("".join(finalSeqList) )
106 Out = open (outputfile, "w") 110 Out = open (outputfile, "w")
107 print >> Out, ">Scaffold" 111 print >> Out, ">Scaffold_from_%s_guided_by_%s" % (prefix, suffix)
108 print >> Out, finalSequence 112 print >> Out, finalSequence
109 Out.close() 113 Out.close()
110 114
111 def __main__(): 115 def __main__():
112 args = Parser() 116 args = Parser()
113 ContigsDict = myContigs (args.sequences) 117 ContigsDict = myContigs (args.sequences)
114 GuideDict = myGuide (args.guideSequence) 118 GuideDict = myGuide (args.guideSequence)
115 blastlist = blatnInfo(args.blast_tab) 119 blastlist = blatnInfo(args.blast_tab)
116 updateGuide(blastlist, GuideDict, ContigsDict) 120 updateGuide(blastlist, GuideDict, ContigsDict)
117 finalAssembly(GuideDict, args.output) 121 finalAssembly(GuideDict, args.output, args.scaffold_prefix, args.scaffold_suffix)
118 122
119 if __name__ == "__main__": 123 if __name__ == "__main__":
120 __main__() 124 __main__()