Mercurial > repos > drosofff > blast_to_scaffold

comparison blast_to_scaffold.py @ 3:e279e53a3fde 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 030207144f0811822dbdda9a10e036ff8e794d7c
author drosofff
date Fri, 25 Mar 2016 19:14:20 -0400
parents 61731b911f4e
children 3288cc5a57e5
comparison
equal deleted inserted replaced
2:61731b911f4e 3:e279e53a3fde
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")
37 args = the_parser.parse_args() 33 args = the_parser.parse_args()
38 return args 34 return args
39 35
40 def blatnInfo (file): 36 def blatnInfo (file):
41 blastlist = [] 37 blastlist = []
100 except KeyError: 96 except KeyError:
101 continue 97 continue
102 for i, nucleotide in enumerate(sequence): 98 for i, nucleotide in enumerate(sequence):
103 GuideDict[i+subjectStart] = nucleotide 99 GuideDict[i+subjectStart] = nucleotide
104 100
105 def finalAssembly (GuideDict, outputfile, prefix, suffix): 101 def finalAssembly (GuideDict, outputfile):
106 finalSeqList = [] 102 finalSeqList = []
107 for keys in sorted(GuideDict): 103 for keys in sorted(GuideDict):
108 finalSeqList.append(GuideDict[keys]) 104 finalSeqList.append(GuideDict[keys])
109 finalSequence = insert_newlines("".join(finalSeqList) ) 105 finalSequence = insert_newlines("".join(finalSeqList) )
110 Out = open (outputfile, "w") 106 Out = open (outputfile, "w")
111 print >> Out, ">Scaffold_from_%s_guided_by_%s" % (prefix, suffix) 107 print >> Out, ">Scaffold"
112 print >> Out, finalSequence 108 print >> Out, finalSequence
113 Out.close() 109 Out.close()
114 110
115 def __main__(): 111 def __main__():
116 args = Parser() 112 args = Parser()
117 ContigsDict = myContigs (args.sequences) 113 ContigsDict = myContigs (args.sequences)
118 GuideDict = myGuide (args.guideSequence) 114 GuideDict = myGuide (args.guideSequence)
119 blastlist = blatnInfo(args.blast_tab) 115 blastlist = blatnInfo(args.blast_tab)
120 updateGuide(blastlist, GuideDict, ContigsDict) 116 updateGuide(blastlist, GuideDict, ContigsDict)
121 finalAssembly(GuideDict, args.output, args.scaffold_prefix, args.scaffold_suffix) 117 finalAssembly(GuideDict, args.output)
122 118
123 if __name__ == "__main__": 119 if __name__ == "__main__":
124 __main__() 120 __main__()