bio-samtools 2.1.0 → 2.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 0a798765b86732cd5e41c96a458439fbc51ec41c
-  data.tar.gz: 43315c48fc5fbc0f182e024031970f7e122b3446
+  metadata.gz: a061cf6e87c05ce1f622d1607878cb68d85f92e0
+  data.tar.gz: 90f4a519c6174d6b8787e42d6be8917f1b5721b1
 SHA512:
-  metadata.gz: abe3c8c927c6459c67fc9a7c79437dad67a9e42c659cd04805de5cc5f8e3b2fc0725a7cb5906e36eb99331b8c3f7e2475d69b532689428b81b40af0361d99b58
-  data.tar.gz: eba5e3fff77e96192f8788eb0cf12d1c34cfd595d9954394d02c3780e45619aeb48065d5dbcea373e0febd6a8be2b179e020b2bbfc014cfa9e831005273908be
+  metadata.gz: 02168c213254e2f4cbe2b16281c0f92c4dbccf6aaca92c4418457a68b835ac2a07e5cc81d68934f86441fe6fb1578ff08e149a5915262843cec67afd5ccc5e8a
+  data.tar.gz: 02b7cb92700fc834a3425649f8c1df6694cb22b3d6bc805707695f7206d1055ebbe72293e9a577df6766341ab0dbf12fa95992eec0addc7b4f8321219c96c979

data/VERSION

@@ -1 +1 @@
-2.1.0
+2.2.0

data/bin/bam_consensus.rb

@@ -0,0 +1,85 @@
+#!/usr/bin/env ruby
+
+require 'bio'
+
+require 'optparse'
+require 'set'
+
+$: << File.expand_path(File.dirname(__FILE__) + '/../lib')
+$: << File.expand_path('.')
+path= File.expand_path(File.dirname(__FILE__) + '/../lib/bio-samtools.rb')
+require path
+
+
+def log(msg)
+  time=Time.now.strftime("%Y-%m-%d %H:%M:%S.%L")
+  puts "#{time}: #{msg}"
+end
+
+
+options = {}
+options[:min_cov] = 5
+options[:min_percentage] = 0.5
+options[:output_file] = "-"
+OptionParser.new do |opts|
+  
+  opts.banner = "Usage: bam_consensus.rb [options]"
+
+  opts.on("-b", "--bam_file FILE", "BAM File to call for the consensus") do |o|
+    options[:bam] = o
+  end
+  opts.on("-r", "--reference FASTA", "reference with the contigs") do |o|
+    options[:reference] = o
+  end
+  opts.on("-p", "--min_percentage FLOAT", "Minimum percentage to call for a base. When more than one base gets the percentage, an ambiguty code is produced") do |o|
+    options[:min_percentage] = o.to_f / 100
+  end
+
+   opts.on("-m", "--min_cov INT", "Minimum percentage to call for a base. When more than one base gets the percentage, an ambiguty code is produced") do |o|
+    options[:min_cov] = o.to_i 
+  end
+
+  opts.on("-f", "--filter_entries FILE", "File with a list of entries to process") do |o|
+    options[:filter_entries] = o
+  end
+
+  opts.on("-o" "--output_file FILE", "Output of the program, in fasta format") do |o|
+    options[:output_file] = o
+  end
+  
+end.parse!
+
+bam =  @parental_1_sam =  Bio::DB::Sam.new({:fasta=>options[:reference], :bam=>options[:bam]})
+region_set = nil
+  if options[:filter_entries]
+  region_set = Set.new
+  File.foreach(options[:filter_entries]) do |line| 
+    region_set << line.chomp
+  end
+end
+
+fasta_db = Bio::DB::Fasta::FastaFile.new(:fasta=> options[:reference])
+fasta_db.load_fai_entries
+
+output = $stdout
+
+output = File.open(options[:output_file], "w") if options[:output_file] != "-"
+
+fasta_db.index.entries.each do | r |
+  process = true
+	region=r.get_full_region
+
+  process = region_set.include? region.entry if region_set
+  if process
+    reg = bam.fetch_region({:region=>region,  :min_cov=>options[:min_cov],:min_per=>options[:min_percentage],  :A=>1})
+    cons = reg.consensus
+    org = fasta_db.fetch_sequence(region)
+    if cons.upcase != org.upcase 
+      output.puts ">#{region.entry}"
+      tmp = cons.scan /.{1,80}/
+      output.puts  tmp.join("\n")
+    end
+  end
+end
+
+output.close if options[:output_file] != "-"

data/bio-samtools.gemspec

@@ -2,19 +2,20 @@
 # DO NOT EDIT THIS FILE DIRECTLY
 # Instead, edit Jeweler::Tasks in Rakefile, and run 'rake gemspec'
 # -*- encoding: utf-8 -*-
-# stub: bio-samtools 2.1.0 ruby lib
+# stub: bio-samtools 2.2.0 ruby lib
 # stub: ext/mkrf_conf.rb
 
 Gem::Specification.new do |s|
   s.name = "bio-samtools"
-  s.version = "2.1.0"
+  s.version = "2.2.0"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.require_paths = ["lib"]
   s.authors = ["Ricardo Ramirez-Gonzalez", "Dan MacLean", "Raoul J.P. Bonnal"]
-  s.date = "2014-09-03"
+  s.date = "2014-10-16"
   s.description = "Binder of samtools for ruby, on the top of FFI. \n\n  This project was born from the need to add support of BAM files to \n  the gee_fu genome browser (http://github.com/danmaclean/gee_fu)."
   s.email = "ilpuccio.febo@gmail.com"
+  s.executables = ["bam_consensus.rb"]
   s.extensions = ["ext/mkrf_conf.rb"]
   s.extra_rdoc_files = [
     "LICENSE.txt",
@@ -28,6 +29,7 @@ Gem::Specification.new do |s|
     "README.md",
     "Rakefile",
     "VERSION",
+    "bin/bam_consensus.rb",
     "bio-samtools.gemspec",
     "doc/Bio.html",
     "doc/Bio/DB.html",

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: bio-samtools
 version: !ruby/object:Gem::Version
-  version: 2.1.0
+  version: 2.2.0
 platform: ruby
 authors:
 - Ricardo Ramirez-Gonzalez
@@ -10,7 +10,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-09-03 00:00:00.000000000 Z
+date: 2014-10-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bio-svgenes
@@ -184,7 +184,8 @@ description: "Binder of samtools for ruby, on the top of FFI. \n\n  This project
   born from the need to add support of BAM files to \n  the gee_fu genome browser
   (http://github.com/danmaclean/gee_fu)."
 email: ilpuccio.febo@gmail.com
-executables: []
+executables:
+- bam_consensus.rb
 extensions:
 - ext/mkrf_conf.rb
 extra_rdoc_files:
@@ -198,6 +199,7 @@ files:
 - README.md
 - Rakefile
 - VERSION
+- bin/bam_consensus.rb
 - bio-samtools.gemspec
 - doc/Bio.html
 - doc/Bio/DB.html