Mercurial > repos > weilong-guo > bs_seeker2

comparison BSseeker2/bs_seeker2-build.py @ 1:8b26adf64adc draft default tip

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
V2.0.5
author weilong-guo
date Tue, 05 Nov 2013 01:55:39 -0500
parents e6df770c0e58
children
comparison
equal deleted inserted replaced
0:e6df770c0e58 1:8b26adf64adc
1 #!/usr/bin/python 1 #!/usr/bin/env python
2 2
3 import os 3 import os
4 from optparse import OptionParser, OptionGroup 4 from optparse import OptionParser, OptionGroup
5 from bs_index.wg_build import * 5 from bs_index.wg_build import *
6 from bs_index.rrbs_build import * 6 from bs_index.rrbs_build import *
10 if __name__ == '__main__': 10 if __name__ == '__main__':
11 11
12 parser = OptionParser() 12 parser = OptionParser()
13 13
14 parser.add_option("-f", "--file", dest="filename", help="Input your reference genome file (fasta)", metavar="FILE") 14 parser.add_option("-f", "--file", dest="filename", help="Input your reference genome file (fasta)", metavar="FILE")
15 parser.add_option("--aligner", dest="aligner", help="Aligner program to perform the analysis: " + ', '.join(supported_aligners) + " [Default: %default]", metavar="ALIGNER", default = BOWTIE2) 15 parser.add_option("--aligner", dest="aligner", help="Aligner program to perform the analysis: " + ', '.join(supported_aligners) + " [Default: %default]", metavar="ALIGNER", default = BOWTIE)
16 parser.add_option("-p", "--path", dest="aligner_path", help="Path to the aligner program. Defaults: " +' '*70+ '\t'.join(('%s: %s '+' '*70) % (al, aligner_path[al]) for al in sorted(supported_aligners)), 16 parser.add_option("-p", "--path", dest="aligner_path", help="Path to the aligner program. Detected: " +' '*70+ '\t'.join(('%s: %s '+' '*70) % (al, aligner_path[al]) for al in sorted(supported_aligners)),
17 metavar="PATH") 17 metavar="PATH")
18 parser.add_option("-d", "--db", type="string", dest="dbpath", help="Path to the reference genome library (generated in preprocessing genome) [Default: %default]", metavar="DBPATH", default = reference_genome_path) 18 parser.add_option("-d", "--db", type="string", dest="dbpath", help="Path to the reference genome library (generated in preprocessing genome) [Default: %default]", metavar="DBPATH", default = reference_genome_path)
19 19
20 parser.add_option("-v", "--version", action="store_true", dest="version", help="show version of BS-Seeker2", default=False) 20 parser.add_option("-v", "--version", action="store_true", dest="version", help="show version of BS-Seeker2", default=False)
21 21
22 # RRBS options 22 # RRBS options
23 rrbs_opts = OptionGroup(parser, "Reduced Representation Bisulfite Sequencing Options", 23 rrbs_opts = OptionGroup(parser, "Reduced Representation Bisulfite Sequencing Options",
24 "Use this options with conjuction of -r [--rrbs]") 24 "Use this options with conjuction of -r [--rrbs]")
25 rrbs_opts.add_option("-r", "--rrbs", action="store_true", dest="rrbs", help = 'Build index specially for Reduced Representation Bisulfite Sequencing experiments. Genome other than certain fragments will be masked. [Default: %default]', default = False) 25 rrbs_opts.add_option("-r", "--rrbs", action="store_true", dest="rrbs", help = 'Build index specially for Reduced Representation Bisulfite Sequencing experiments. Genome other than certain fragments will be masked. [Default: %default]', default = False)
26 rrbs_opts.add_option("-l", "--low",type= "int", dest="low_bound", help="lower bound of fragment length (excluding recognition sequence such as C-CGG) [Default: %default]", default = 40) 26 rrbs_opts.add_option("-l", "--low",type= "int", dest="low_bound", help="lower bound of fragment length (excluding recognition sequence such as C-CGG) [Default: %default]", default = 20)
27 rrbs_opts.add_option("-u", "--up", type= "int", dest="up_bound", help="upper bound of fragment length (excluding recognition sequence such as C-CGG ends) [Default: %default]", default = 500) 27 rrbs_opts.add_option("-u", "--up", type= "int", dest="up_bound", help="upper bound of fragment length (excluding recognition sequence such as C-CGG ends) [Default: %default]", default = 500)
28 rrbs_opts.add_option("-c", "--cut-site", type= "string", dest="cut_format", help="Cut sites of restriction enzyme. Ex: MspI(C-CGG), Mael:(C-TAG), double-enzyme MspI&Mael:(C-CGG,C-TAG). [Default: %default]", default = "C-CGG") 28 rrbs_opts.add_option("-c", "--cut-site", type= "string", dest="cut_format", help="Cut sites of restriction enzyme. Ex: MspI(C-CGG), Mael:(C-TAG), double-enzyme MspI&Mael:(C-CGG,C-TAG). [Default: %default]", default = "C-CGG")
29 parser.add_option_group(rrbs_opts) 29 parser.add_option_group(rrbs_opts)
30 30
31 31
32 (options, args) = parser.parse_args() 32 (options, args) = parser.parse_args()
33 33
34 # if no options were given by the user, print help and exit 34 # if no options were given by the user, print help and exit
35 if len(sys.argv) == 1: 35 if len(sys.argv) == 1:
36 print parser.print_help() 36 parser.print_help()
37 exit(0) 37 exit(0)
38 38
39 if options.version : 39 if options.version :
40 show_version() 40 show_version()
41 exit (-1) 41 exit (-1)
42 else : 42 else :
43 show_version() 43 show_version()
44 44
45 rrbs = options.rrbs 45 rrbs = options.rrbs
46 46
47 fasta_file=os.path.expanduser(options.filename) 47 if options.filename is not None :
48 fasta_file=os.path.expanduser(options.filename)
49 else :
50 error("Please specify the genome file (Fasta) using \"-f\"")
51
48 if fasta_file is None: 52 if fasta_file is None:
49 error('Fasta file for the reference genome must be supported') 53 error('Fasta file for the reference genome must be supported')
50 54
51 if not os.path.isfile(fasta_file): 55 if not os.path.isfile(fasta_file):
52 error('%s cannot be found' % fasta_file) 56 error('%s cannot be found' % fasta_file)
67 }[options.aligner] 71 }[options.aligner]
68 72
69 73
70 print "Reference genome file: %s" % fasta_file 74 print "Reference genome file: %s" % fasta_file
71 print "Reduced Representation Bisulfite Sequencing: %s" % rrbs 75 print "Reduced Representation Bisulfite Sequencing: %s" % rrbs
76 print "Short reads aligner you are using: %s" % options.aligner
72 print "Builder path: %s" % builder_exec 77 print "Builder path: %s" % builder_exec
78
73 #--------------------------------------------------------------- 79 #---------------------------------------------------------------
80
81 if not os.path.isfile( builder_exec ) :
82 error("Cannot file program %s for execution." % builder_exec)
74 83
75 ref_path = options.dbpath 84 ref_path = options.dbpath
76 85
77 if os.path.exists(ref_path): 86 if os.path.exists(ref_path):
78 if not os.path.isdir(ref_path): 87 if not os.path.isdir(ref_path):