bio-genomic-interval 0.1.0
Sign up to get free protection for your applications and to get access to all the features.
- data/.document +5 -0
- data/.rspec +2 -0
- data/Gemfile +14 -0
- data/Gemfile.lock +30 -0
- data/LICENSE.txt +20 -0
- data/README.rdoc +79 -0
- data/Rakefile +51 -0
- data/VERSION +1 -0
- data/bio-genomic-interval.gemspec +67 -0
- data/lib/bio-genomic-interval.rb +158 -0
- data/spec/bio-genomic-interval_spec.rb +357 -0
- data/spec/spec_helper.rb +12 -0
- metadata +157 -0
data/.document
ADDED
data/.rspec
ADDED
data/Gemfile
ADDED
@@ -0,0 +1,14 @@
|
|
1
|
+
source "http://rubygems.org"
|
2
|
+
# Add dependencies required to use your gem here.
|
3
|
+
# Example:
|
4
|
+
# gem "activesupport", ">= 2.3.5"
|
5
|
+
|
6
|
+
# Add dependencies to develop your gem here.
|
7
|
+
# Include everything needed to run rake, tests, features, etc.
|
8
|
+
group :development do
|
9
|
+
gem "rspec", "~> 2.3.0"
|
10
|
+
gem "bundler", "~> 1.0.0"
|
11
|
+
gem "jeweler", "~> 1.5.2"
|
12
|
+
gem "rcov", ">= 0"
|
13
|
+
gem "bio", ">= 1.4.1"
|
14
|
+
end
|
data/Gemfile.lock
ADDED
@@ -0,0 +1,30 @@
|
|
1
|
+
GEM
|
2
|
+
remote: http://rubygems.org/
|
3
|
+
specs:
|
4
|
+
bio (1.4.1)
|
5
|
+
diff-lcs (1.1.2)
|
6
|
+
git (1.2.5)
|
7
|
+
jeweler (1.5.2)
|
8
|
+
bundler (~> 1.0.0)
|
9
|
+
git (>= 1.2.5)
|
10
|
+
rake
|
11
|
+
rake (0.8.7)
|
12
|
+
rcov (0.9.9)
|
13
|
+
rspec (2.3.0)
|
14
|
+
rspec-core (~> 2.3.0)
|
15
|
+
rspec-expectations (~> 2.3.0)
|
16
|
+
rspec-mocks (~> 2.3.0)
|
17
|
+
rspec-core (2.3.1)
|
18
|
+
rspec-expectations (2.3.0)
|
19
|
+
diff-lcs (~> 1.1.2)
|
20
|
+
rspec-mocks (2.3.0)
|
21
|
+
|
22
|
+
PLATFORMS
|
23
|
+
ruby
|
24
|
+
|
25
|
+
DEPENDENCIES
|
26
|
+
bio (>= 1.4.1)
|
27
|
+
bundler (~> 1.0.0)
|
28
|
+
jeweler (~> 1.5.2)
|
29
|
+
rcov
|
30
|
+
rspec (~> 2.3.0)
|
data/LICENSE.txt
ADDED
@@ -0,0 +1,20 @@
|
|
1
|
+
Copyright (c) 2011 Hiroyuki Mishima
|
2
|
+
|
3
|
+
Permission is hereby granted, free of charge, to any person obtaining
|
4
|
+
a copy of this software and associated documentation files (the
|
5
|
+
"Software"), to deal in the Software without restriction, including
|
6
|
+
without limitation the rights to use, copy, modify, merge, publish,
|
7
|
+
distribute, sublicense, and/or sell copies of the Software, and to
|
8
|
+
permit persons to whom the Software is furnished to do so, subject to
|
9
|
+
the following conditions:
|
10
|
+
|
11
|
+
The above copyright notice and this permission notice shall be
|
12
|
+
included in all copies or substantial portions of the Software.
|
13
|
+
|
14
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
|
15
|
+
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
|
16
|
+
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
|
17
|
+
NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
|
18
|
+
LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
|
19
|
+
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
|
20
|
+
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
|
data/README.rdoc
ADDED
@@ -0,0 +1,79 @@
|
|
1
|
+
= bio-genomic-interval
|
2
|
+
|
3
|
+
Author:: MISHIMA, Hiroyuki (hmishima AT nagasaki-u.ac.jp, missy AT be.to)
|
4
|
+
Version:: 0.1.0
|
5
|
+
Copyright:: Copyright (c) MISHIMA, Hiroyuki, 2011
|
6
|
+
License:: the MIT/X11 license
|
7
|
+
|
8
|
+
A BioRuby plugin: handling genomic intervals,such as "chr1:123-456",
|
9
|
+
and overlaps between two intervals.
|
10
|
+
|
11
|
+
= Usage
|
12
|
+
== Generation of interval objects
|
13
|
+
a interval object be generated by like the following:
|
14
|
+
a = Bio::GenomicInterval.new("chr1", 123, 456)
|
15
|
+
a.to_s # => "chr1:123-456"
|
16
|
+
b = Bio::GenomicInterval.parse("chr1:123-456")
|
17
|
+
b.to_s # => "chr1:123-456"
|
18
|
+
c = Bio::GenomicInetrval.zero_based("chr1", 122, 456)
|
19
|
+
c.to_s # => "chr1:123-456"
|
20
|
+
c.zero_start # => 122
|
21
|
+
|
22
|
+
Last one is generated from a "Zero-based half-closed[start, end)" interval
|
23
|
+
, which used in UCSC Genobe Browser's BED format, instead of usual
|
24
|
+
"One-based full-closed [start, end]" intervals.
|
25
|
+
|
26
|
+
== Comparison
|
27
|
+
ref = Bio::GenomicInterval.parse("chr1:123-456")
|
28
|
+
cmp = Bio::GenomicInterval.parse("chr1:234-567")
|
29
|
+
ref.compare(cmp) # => :right_overlapped
|
30
|
+
|
31
|
+
ref.adjacent # => 20
|
32
|
+
near = Bio::GenomicInterval.parse("chr1:458-567")
|
33
|
+
ref.compare(cmp) # => :right_adjacent
|
34
|
+
|
35
|
+
ref.adjacent = 1
|
36
|
+
ref.compare(cmp) # => :right_off
|
37
|
+
|
38
|
+
== Overlap metrics
|
39
|
+
* When a overlap exist, return a positive integers (>1) for the overlap length.
|
40
|
+
* When a overlap does not exist, return a zero or a negative (<= 0) for the space size between the intervals.
|
41
|
+
|
42
|
+
ref = Bio::GenomicInterval.parse("chr1:10-20")
|
43
|
+
cmp = Bio::GenomicInterval.parse("chr1:15-25")
|
44
|
+
ref.overlap(cmp) # => 6
|
45
|
+
cmp2 = Bio::GenomicInterval.parse("chr1:25-35")
|
46
|
+
ref.overlap(cmp) # => -4
|
47
|
+
|
48
|
+
== Expansion (or integration)
|
49
|
+
ref = Bio::GenomicInterval.parse("chr1:400-600")
|
50
|
+
other = Bio::GenomicInterval.parse("chr1:650-800")
|
51
|
+
ref.expand(other).to_s # => "chr1:400-800"
|
52
|
+
|
53
|
+
== Center
|
54
|
+
obj1 = Bio::GenomicInterval.parse("chr1:1-3")
|
55
|
+
obj1.center # => "chr1:2-2"
|
56
|
+
obj2 = Bio::GenomicInterval.parse("chr2:1-4")
|
57
|
+
obj2.center # => "chr1:2-2"
|
58
|
+
|
59
|
+
== And others
|
60
|
+
ref = Bio::GenomicInterval.parse("chr1:400-600")
|
61
|
+
other = Bio::GenomicInterval.parse("chr1:605-800")
|
62
|
+
ref.overlapped?(other) # => false
|
63
|
+
ref.nearly_overlapped?(other) # => true
|
64
|
+
ref.size # => 201
|
65
|
+
ref.chr_start -= 100
|
66
|
+
ref.chr_end += 100
|
67
|
+
ref.chrom = "chrX"
|
68
|
+
ref.to_s # => "chrX:300-700"
|
69
|
+
|
70
|
+
See also the Rspec file.
|
71
|
+
|
72
|
+
== Contributing to bio-genomic-interval
|
73
|
+
Please do not hesitate to contanct the author by emails.
|
74
|
+
|
75
|
+
== Copyright
|
76
|
+
|
77
|
+
Copyright (c) 2011 Hiroyuki Mishima. See LICENSE.txt for
|
78
|
+
further details.
|
79
|
+
|
data/Rakefile
ADDED
@@ -0,0 +1,51 @@
|
|
1
|
+
require 'rubygems'
|
2
|
+
require 'bundler'
|
3
|
+
begin
|
4
|
+
Bundler.setup(:default, :development)
|
5
|
+
rescue Bundler::BundlerError => e
|
6
|
+
$stderr.puts e.message
|
7
|
+
$stderr.puts "Run `bundle install` to install missing gems"
|
8
|
+
exit e.status_code
|
9
|
+
end
|
10
|
+
require 'rake'
|
11
|
+
|
12
|
+
require 'jeweler'
|
13
|
+
Jeweler::Tasks.new do |gem|
|
14
|
+
# gem is a Gem::Specification... see http://docs.rubygems.org/read/chapter/20 for more options
|
15
|
+
gem.name = "bio-genomic-interval"
|
16
|
+
gem.homepage = "http://github.com/misshie/bioruby-genomic-interval"
|
17
|
+
gem.license = "MIT"
|
18
|
+
gem.summary = %Q{a BioRuby plugin: handling genomic interavals and overlaps}
|
19
|
+
gem.description =
|
20
|
+
%Q{a BioRuby plugin: handling genomic intervals,such as "chr1:123-456", and overlaps between two intervalsrake }
|
21
|
+
gem.email = "missy@be.to"
|
22
|
+
gem.authors = ["Hiroyuki Mishima"]
|
23
|
+
# Include your dependencies below. Runtime dependencies are required when using your gem,
|
24
|
+
# and development dependencies are only needed for development (ie running rake tasks, tests, etc)
|
25
|
+
# gem.add_runtime_dependency 'jabber4r', '> 0.1'
|
26
|
+
# gem.add_development_dependency 'rspec', '> 1.2.3'
|
27
|
+
end
|
28
|
+
Jeweler::RubygemsDotOrgTasks.new
|
29
|
+
|
30
|
+
require 'rspec/core'
|
31
|
+
require 'rspec/core/rake_task'
|
32
|
+
RSpec::Core::RakeTask.new(:spec) do |spec|
|
33
|
+
spec.pattern = FileList['spec/**/*_spec.rb']
|
34
|
+
end
|
35
|
+
|
36
|
+
RSpec::Core::RakeTask.new(:rcov) do |spec|
|
37
|
+
spec.pattern = 'spec/**/*_spec.rb'
|
38
|
+
spec.rcov = true
|
39
|
+
end
|
40
|
+
|
41
|
+
task :default => :spec
|
42
|
+
|
43
|
+
require 'rake/rdoctask'
|
44
|
+
Rake::RDocTask.new do |rdoc|
|
45
|
+
version = File.exist?('VERSION') ? File.read('VERSION') : ""
|
46
|
+
|
47
|
+
rdoc.rdoc_dir = 'rdoc'
|
48
|
+
rdoc.title = "bio-genomic-interval #{version}"
|
49
|
+
rdoc.rdoc_files.include('README*')
|
50
|
+
rdoc.rdoc_files.include('lib/**/*.rb')
|
51
|
+
end
|
data/VERSION
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
0.1.0
|
@@ -0,0 +1,67 @@
|
|
1
|
+
# Generated by jeweler
|
2
|
+
# DO NOT EDIT THIS FILE DIRECTLY
|
3
|
+
# Instead, edit Jeweler::Tasks in Rakefile, and run 'rake gemspec'
|
4
|
+
# -*- encoding: utf-8 -*-
|
5
|
+
|
6
|
+
Gem::Specification.new do |s|
|
7
|
+
s.name = %q{bio-genomic-interval}
|
8
|
+
s.version = "0.1.0"
|
9
|
+
|
10
|
+
s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
|
11
|
+
s.authors = ["Hiroyuki Mishima"]
|
12
|
+
s.date = %q{2011-04-07}
|
13
|
+
s.description = %q{a BioRuby plugin: handling genomic intervals,such as "chr1:123-456", and overlaps between two intervalsrake }
|
14
|
+
s.email = %q{missy@be.to}
|
15
|
+
s.extra_rdoc_files = [
|
16
|
+
"LICENSE.txt",
|
17
|
+
"README.rdoc"
|
18
|
+
]
|
19
|
+
s.files = [
|
20
|
+
".document",
|
21
|
+
".rspec",
|
22
|
+
"Gemfile",
|
23
|
+
"Gemfile.lock",
|
24
|
+
"LICENSE.txt",
|
25
|
+
"README.rdoc",
|
26
|
+
"Rakefile",
|
27
|
+
"VERSION",
|
28
|
+
"bio-genomic-interval.gemspec",
|
29
|
+
"lib/bio-genomic-interval.rb",
|
30
|
+
"spec/bio-genomic-interval_spec.rb",
|
31
|
+
"spec/spec_helper.rb"
|
32
|
+
]
|
33
|
+
s.homepage = %q{http://github.com/misshie/bioruby-genomic-interval}
|
34
|
+
s.licenses = ["MIT"]
|
35
|
+
s.require_paths = ["lib"]
|
36
|
+
s.rubygems_version = %q{1.6.2}
|
37
|
+
s.summary = %q{a BioRuby plugin: handling genomic interavals and overlaps}
|
38
|
+
s.test_files = [
|
39
|
+
"spec/bio-genomic-interval_spec.rb",
|
40
|
+
"spec/spec_helper.rb"
|
41
|
+
]
|
42
|
+
|
43
|
+
if s.respond_to? :specification_version then
|
44
|
+
s.specification_version = 3
|
45
|
+
|
46
|
+
if Gem::Version.new(Gem::VERSION) >= Gem::Version.new('1.2.0') then
|
47
|
+
s.add_development_dependency(%q<rspec>, ["~> 2.3.0"])
|
48
|
+
s.add_development_dependency(%q<bundler>, ["~> 1.0.0"])
|
49
|
+
s.add_development_dependency(%q<jeweler>, ["~> 1.5.2"])
|
50
|
+
s.add_development_dependency(%q<rcov>, [">= 0"])
|
51
|
+
s.add_development_dependency(%q<bio>, [">= 1.4.1"])
|
52
|
+
else
|
53
|
+
s.add_dependency(%q<rspec>, ["~> 2.3.0"])
|
54
|
+
s.add_dependency(%q<bundler>, ["~> 1.0.0"])
|
55
|
+
s.add_dependency(%q<jeweler>, ["~> 1.5.2"])
|
56
|
+
s.add_dependency(%q<rcov>, [">= 0"])
|
57
|
+
s.add_dependency(%q<bio>, [">= 1.4.1"])
|
58
|
+
end
|
59
|
+
else
|
60
|
+
s.add_dependency(%q<rspec>, ["~> 2.3.0"])
|
61
|
+
s.add_dependency(%q<bundler>, ["~> 1.0.0"])
|
62
|
+
s.add_dependency(%q<jeweler>, ["~> 1.5.2"])
|
63
|
+
s.add_dependency(%q<rcov>, [">= 0"])
|
64
|
+
s.add_dependency(%q<bio>, [">= 1.4.1"])
|
65
|
+
end
|
66
|
+
end
|
67
|
+
|
@@ -0,0 +1,158 @@
|
|
1
|
+
module Bio
|
2
|
+
|
3
|
+
#
|
4
|
+
# =a class for manupirate genomic intervals such as "chr1:123-456"
|
5
|
+
#
|
6
|
+
class GenomicInterval
|
7
|
+
# a default value to determine the adjacent/very near distance in bp
|
8
|
+
DEFAULT_ADJACENT = 20
|
9
|
+
|
10
|
+
def initialize(chrom = "", chr_start = 1, chr_end = 1)
|
11
|
+
raise ArgumentError unless chr_start >= 1
|
12
|
+
raise ArgumentError unless chr_end >= 1
|
13
|
+
raise ArgumentError unless chr_start <= chr_end
|
14
|
+
@chrom = chrom
|
15
|
+
@chr_start = chr_start
|
16
|
+
@chr_end = chr_end
|
17
|
+
@adjacent = DEFAULT_ADJACENT
|
18
|
+
end
|
19
|
+
|
20
|
+
attr_accessor :chrom, :chr_start, :chr_end, :adjacent
|
21
|
+
|
22
|
+
#
|
23
|
+
# generate an interval object from a string expressing
|
24
|
+
# one-based full-closed interval such as "chr1:123-456"
|
25
|
+
#
|
26
|
+
def self.parse(interval)
|
27
|
+
chrom, start_end = interval.split(":")
|
28
|
+
str_start, str_end = start_end.split("-")[0..1]
|
29
|
+
str_end = str_start if str_end.nil?
|
30
|
+
chr_start = Integer(str_start)
|
31
|
+
chr_end = Integer(str_end)
|
32
|
+
if chr_start > chr_end
|
33
|
+
chr_end, chr_start = chr_start, chr_end
|
34
|
+
end
|
35
|
+
self.new(chrom, chr_start, chr_end)
|
36
|
+
end
|
37
|
+
|
38
|
+
#
|
39
|
+
# generate an interval object from three atguments expressing
|
40
|
+
# zero-based half-closed interval such as "chr1", 122, 456
|
41
|
+
#
|
42
|
+
def self.zero_based(chrom = "", z_start = 0, z_end = 1)
|
43
|
+
z_start += 1
|
44
|
+
self.new(chrom, z_start, z_end)
|
45
|
+
end
|
46
|
+
|
47
|
+
#
|
48
|
+
# returns one-based full-closed interval such as "chr1:123-456"
|
49
|
+
#
|
50
|
+
def to_s
|
51
|
+
"#{@chrom}:#{@chr_start}-#{@chr_end}"
|
52
|
+
end
|
53
|
+
|
54
|
+
#
|
55
|
+
# returns zero-based half-closed start position
|
56
|
+
#
|
57
|
+
def zero_start
|
58
|
+
@chr_start - 1
|
59
|
+
end
|
60
|
+
|
61
|
+
#
|
62
|
+
# returns zero-based half-closed end position
|
63
|
+
#
|
64
|
+
def zero_end
|
65
|
+
@chr_end
|
66
|
+
end
|
67
|
+
|
68
|
+
#
|
69
|
+
# returns one of the followings:
|
70
|
+
# :different_chrom, :left_adjacent, :right_adjacent
|
71
|
+
# :left_off, :right_off, :equal
|
72
|
+
# :contained, :containing, :left_overlapped, :right_overlapped
|
73
|
+
# Imagine that the receiver object is fixed on a number line
|
74
|
+
#
|
75
|
+
def compare(other)
|
76
|
+
case
|
77
|
+
when self.chrom != other.chrom
|
78
|
+
:different_chrom
|
79
|
+
when other.chr_end.between?(self.chr_start - @adjacent, self.chr_start - 1)
|
80
|
+
:left_adjacent
|
81
|
+
when other.chr_start.between?(self.chr_end + 1, self.chr_end + @adjacent)
|
82
|
+
:right_adjacent
|
83
|
+
when other.chr_end < self.chr_start
|
84
|
+
:left_off
|
85
|
+
when self.chr_end < other.chr_start
|
86
|
+
:right_off
|
87
|
+
when (self.chr_start == other.chr_start) &&
|
88
|
+
(self.chr_end == other.chr_end)
|
89
|
+
:equal
|
90
|
+
when (other.chr_start.between?(self.chr_start, self.chr_end)) &&
|
91
|
+
(other.chr_end.between?(self.chr_start, self.chr_end))
|
92
|
+
:contained
|
93
|
+
when (self.chr_start.between?(other.chr_start, other.chr_end)) &&
|
94
|
+
(self.chr_end.between?(other.chr_start, other.chr_end))
|
95
|
+
:containing
|
96
|
+
when (other.chr_start < self.chr_start) &&
|
97
|
+
(other.chr_end.between?(self.chr_start, self.chr_end))
|
98
|
+
:left_overlapped
|
99
|
+
when (other.chr_start.between?(self.chr_start, self.chr_end)) &&
|
100
|
+
(self.chr_end < other.chr_end)
|
101
|
+
:right_overlapped
|
102
|
+
else
|
103
|
+
raise Exception, "must not happen"
|
104
|
+
end
|
105
|
+
end
|
106
|
+
|
107
|
+
def nearly_overlapped?(other)
|
108
|
+
result = compare(other)
|
109
|
+
[ :left_adjacent, :right_adjacent,
|
110
|
+
:equal, :contained, :containing,
|
111
|
+
:left_overlapped, :right_overlapped].any?{|x| x == result}
|
112
|
+
end
|
113
|
+
|
114
|
+
def overlapped?(other)
|
115
|
+
result = compare(other)
|
116
|
+
[ :equal, :contained, :containing,
|
117
|
+
:left_overlapped, :right_overlapped].any?{|x| x == result}
|
118
|
+
end
|
119
|
+
|
120
|
+
def expand(other)
|
121
|
+
raise ArgumentError unless self.chrom == other.chrom
|
122
|
+
new_start = [self.chr_start, other.chr_start].min
|
123
|
+
new_end = [self.chr_end, other.chr_end].max
|
124
|
+
Bio::GenomicInterval.new(@chrom, new_start, new_end)
|
125
|
+
end
|
126
|
+
|
127
|
+
def size
|
128
|
+
chr_end - chr_start + 1
|
129
|
+
end
|
130
|
+
|
131
|
+
alias :length :size
|
132
|
+
|
133
|
+
def center
|
134
|
+
center = (chr_start + chr_end) / 2
|
135
|
+
Bio::GenomicInterval.new(self.chrom, center, center)
|
136
|
+
end
|
137
|
+
|
138
|
+
# * When a overlap exist, return a positive integers (>1) for the overlap length.
|
139
|
+
# * When a overlap does not exist, return a zero or a negative (<= 0) for the space size between the intervals.
|
140
|
+
#
|
141
|
+
def overlap(other)
|
142
|
+
case self.compare(other)
|
143
|
+
when :different_chrom
|
144
|
+
0
|
145
|
+
when :left_off, :left_adjacent, :left_overlapped
|
146
|
+
other.chr_end - self.chr_start + 1
|
147
|
+
when :contained, :equal
|
148
|
+
other.size
|
149
|
+
when :containing
|
150
|
+
self.size
|
151
|
+
when :right_off, :right_adjacent, :right_overlapped
|
152
|
+
self.chr_end - other.chr_start + 1
|
153
|
+
else
|
154
|
+
raise Exception, "must not happen"
|
155
|
+
end
|
156
|
+
end
|
157
|
+
end
|
158
|
+
end
|
@@ -0,0 +1,357 @@
|
|
1
|
+
require File.expand_path(File.dirname(__FILE__) + '/spec_helper')
|
2
|
+
|
3
|
+
describe "Bio::GenomicInterval" do
|
4
|
+
|
5
|
+
#
|
6
|
+
# class methods
|
7
|
+
#
|
8
|
+
|
9
|
+
describe ".new" do
|
10
|
+
context 'given ("chr1", 1, 234)' do
|
11
|
+
it "is a Bio::GenomicInterval" do
|
12
|
+
Bio::GenomicInterval.new("chr1", 1, 234).should be_a(Bio::GenomicInterval)
|
13
|
+
end
|
14
|
+
end
|
15
|
+
|
16
|
+
context 'given ("chr1", -1, 234)' do
|
17
|
+
it "raises an ArgumentError" do
|
18
|
+
expect{Bio::GenomicInterval.new("chr1", -1, 234)}.to raise_error(ArgumentError)
|
19
|
+
end
|
20
|
+
end
|
21
|
+
|
22
|
+
context 'given ("chr1", 234, 1)' do
|
23
|
+
it "raise an ArgumentError" do
|
24
|
+
expect{Bio::GenomicInterval.new("chr1", 234, 1)}.to raise_error(ArgumentError)
|
25
|
+
end
|
26
|
+
end
|
27
|
+
|
28
|
+
context 'given no argument' do
|
29
|
+
it "is a Bio::GenomicInterval" do
|
30
|
+
Bio::GenomicInterval.new.should be_a(Bio::GenomicInterval)
|
31
|
+
end
|
32
|
+
end
|
33
|
+
end
|
34
|
+
|
35
|
+
describe ".parse" do
|
36
|
+
context 'given "chr1:1-123"' do
|
37
|
+
it 'represents "chr1:1-123" by the to_s method' do
|
38
|
+
Bio::GenomicInterval.parse("chr1:1-123").to_s.should == "chr1:1-123"
|
39
|
+
end
|
40
|
+
end
|
41
|
+
|
42
|
+
context 'given "chr1:123-1"' do
|
43
|
+
it 'represents "chr1:1-123" by the to_s method' do
|
44
|
+
Bio::GenomicInterval.parse("chr1:123-1").to_s.should == "chr1:1-123"
|
45
|
+
end
|
46
|
+
end
|
47
|
+
|
48
|
+
context 'given "chr1:123"' do
|
49
|
+
it 'represents "chr1:123-123" by the to_s method' do
|
50
|
+
Bio::GenomicInterval.parse("chr1:123").to_s.should == "chr1:123-123"
|
51
|
+
end
|
52
|
+
end
|
53
|
+
end
|
54
|
+
|
55
|
+
describe ".zero_based" do
|
56
|
+
context 'given ("chr1", 0, 1)' do
|
57
|
+
it 'represents "chr1:1-1" by the to_s method' do
|
58
|
+
Bio::GenomicInterval.zero_based("chr1", 0, 1).to_s.should == "chr1:1-1"
|
59
|
+
end
|
60
|
+
end
|
61
|
+
end
|
62
|
+
|
63
|
+
#
|
64
|
+
# instance methods
|
65
|
+
#
|
66
|
+
describe '#zero_start for "chr1:1-1"' do
|
67
|
+
context 'when called' do
|
68
|
+
it 'returns 0' do
|
69
|
+
Bio::GenomicInterval.parse("chr1:1-1").zero_start.should == 0
|
70
|
+
end
|
71
|
+
end
|
72
|
+
end
|
73
|
+
|
74
|
+
describe '#zero_end for "chr1:1-1"' do
|
75
|
+
context 'when called' do
|
76
|
+
it 'returns 1' do
|
77
|
+
Bio::GenomicInterval.parse("chr1:1-1").zero_end.should == 1
|
78
|
+
end
|
79
|
+
end
|
80
|
+
end
|
81
|
+
|
82
|
+
|
83
|
+
describe '#adjacent' do
|
84
|
+
context 'when called first' do
|
85
|
+
it 'returns default value' do
|
86
|
+
default = Bio::GenomicInterval::DEFAULT_ADJACENT
|
87
|
+
Bio::GenomicInterval.new.adjacent.should == default
|
88
|
+
end
|
89
|
+
end
|
90
|
+
|
91
|
+
context 'when set by #adjacent = 10' do
|
92
|
+
it 'returns 10' do
|
93
|
+
obj = Bio::GenomicInterval.new
|
94
|
+
obj.adjacent = 10
|
95
|
+
obj.adjacent.should == 10
|
96
|
+
end
|
97
|
+
end
|
98
|
+
end
|
99
|
+
|
100
|
+
describe '#compare for "chr1:400-600"' do
|
101
|
+
context 'given "chrX:123-234"' do
|
102
|
+
it 'returns :different_chr' do
|
103
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
104
|
+
subject = Bio::GenomicInterval.parse("chrX:123-234")
|
105
|
+
receiver.compare(subject).should == :different_chrom
|
106
|
+
end
|
107
|
+
end
|
108
|
+
|
109
|
+
context 'given "chr1:123-234"' do
|
110
|
+
it 'returns :left_off' do
|
111
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
112
|
+
subject = Bio::GenomicInterval.parse("chr1:123-234")
|
113
|
+
receiver.compare(subject).should == :left_off
|
114
|
+
end
|
115
|
+
end
|
116
|
+
|
117
|
+
context 'given "chr1:789-890"' do
|
118
|
+
it 'returns :right_off' do
|
119
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
120
|
+
subject = Bio::GenomicInterval.parse("chr1:789-890")
|
121
|
+
receiver.compare(subject).should == :right_off
|
122
|
+
end
|
123
|
+
end
|
124
|
+
|
125
|
+
context 'given "chr1:450-550"' do
|
126
|
+
it 'returns :contained' do
|
127
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
128
|
+
subject = Bio::GenomicInterval.parse("chr1:450-550")
|
129
|
+
receiver.compare(subject).should == :contained
|
130
|
+
end
|
131
|
+
end
|
132
|
+
|
133
|
+
context 'given "chr1:400-400" (size == 1)' do
|
134
|
+
it 'returns :contained' do
|
135
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
136
|
+
subject = Bio::GenomicInterval.parse("chr1:400-400")
|
137
|
+
receiver.compare(subject).should == :contained
|
138
|
+
end
|
139
|
+
end
|
140
|
+
|
141
|
+
context 'given "chr1:300-700"' do
|
142
|
+
it 'returns :containing' do
|
143
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
144
|
+
subject = Bio::GenomicInterval.parse("chr1:300-700")
|
145
|
+
receiver.compare(subject).should == :containing
|
146
|
+
end
|
147
|
+
end
|
148
|
+
|
149
|
+
context 'given "chr1:300-500"' do
|
150
|
+
it 'returns :left_overlapped' do
|
151
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
152
|
+
subject = Bio::GenomicInterval.parse("chr1:300-500")
|
153
|
+
receiver.compare(subject).should == :left_overlapped
|
154
|
+
end
|
155
|
+
end
|
156
|
+
|
157
|
+
context 'given "chr1:500-700"' do
|
158
|
+
it 'returns :right_overlapped' do
|
159
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
160
|
+
subject = Bio::GenomicInterval.parse("chr1:500-700")
|
161
|
+
receiver.compare(subject).should == :right_overlapped
|
162
|
+
end
|
163
|
+
end
|
164
|
+
|
165
|
+
context 'given same interval' do
|
166
|
+
it 'returns :equal' do
|
167
|
+
receiver = subject = Bio::GenomicInterval.parse("chr1:400-600")
|
168
|
+
receiver.compare(subject).should == :equal
|
169
|
+
end
|
170
|
+
end
|
171
|
+
|
172
|
+
context 'given "chr1:300-398"' do
|
173
|
+
it 'returns :right_adjacent' do
|
174
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
175
|
+
subject = Bio::GenomicInterval.parse("chr1:300-398")
|
176
|
+
receiver.compare(subject).should == :left_adjacent
|
177
|
+
end
|
178
|
+
end
|
179
|
+
|
180
|
+
context 'given "chr1:399-399" (size == 1)' do
|
181
|
+
it 'returns :right_adjacent' do
|
182
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
183
|
+
subject = Bio::GenomicInterval.parse("chr1:399-399")
|
184
|
+
receiver.compare(subject).should == :left_adjacent
|
185
|
+
end
|
186
|
+
end
|
187
|
+
|
188
|
+
context 'given "chr1:603-800"' do
|
189
|
+
it 'returns :right_adjacent' do
|
190
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
191
|
+
subject = Bio::GenomicInterval.parse("chr1:603-800")
|
192
|
+
receiver.compare(subject).should == :right_adjacent
|
193
|
+
end
|
194
|
+
end
|
195
|
+
end
|
196
|
+
|
197
|
+
describe '#nearly_overlapped? for "chr1:400-600"' do
|
198
|
+
context 'given "chr1:300-500"' do
|
199
|
+
it 'returens true' do
|
200
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
201
|
+
subject = Bio::GenomicInterval.parse("chr1:300-500")
|
202
|
+
receiver.nearly_overlapped?(subject).should be_true
|
203
|
+
end
|
204
|
+
end
|
205
|
+
|
206
|
+
context 'given "chr1:300-390"' do
|
207
|
+
it 'returens true' do
|
208
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
209
|
+
receiver.adjacent = 20
|
210
|
+
subject = Bio::GenomicInterval.parse("chr1:300-390")
|
211
|
+
receiver.nearly_overlapped?(subject).should be_true
|
212
|
+
end
|
213
|
+
end
|
214
|
+
end
|
215
|
+
|
216
|
+
describe '#overlapped? for "chr1:400-600"' do
|
217
|
+
context 'given "chr1:300-500"' do
|
218
|
+
it 'returens true' do
|
219
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
220
|
+
subject = Bio::GenomicInterval.parse("chr1:300-500")
|
221
|
+
receiver.overlapped?(subject).should be_true
|
222
|
+
end
|
223
|
+
end
|
224
|
+
context 'given "chr1:300-370"' do
|
225
|
+
it 'returens false' do
|
226
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
227
|
+
receiver.adjacent = 20
|
228
|
+
subject = Bio::GenomicInterval.parse("chr1:300-370")
|
229
|
+
receiver.overlapped?(subject).should be_false
|
230
|
+
end
|
231
|
+
end
|
232
|
+
end
|
233
|
+
|
234
|
+
describe '#expand for "chr1:400-600"' do
|
235
|
+
context 'given "chr1:603-800"' do
|
236
|
+
it 'returns "chr1:400-800" by the to_s method' do
|
237
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
238
|
+
subject = Bio::GenomicInterval.parse("chr1:603-800")
|
239
|
+
receiver.expand(subject).to_s.should == "chr1:400-800"
|
240
|
+
end
|
241
|
+
end
|
242
|
+
|
243
|
+
context 'given "chrX:603-800"' do
|
244
|
+
it 'raises ArgumentError' do
|
245
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-600")
|
246
|
+
subject = Bio::GenomicInterval.parse("chrX:603-800")
|
247
|
+
expect{receiver.expand(subject)}.to raise_error(ArgumentError)
|
248
|
+
end
|
249
|
+
end
|
250
|
+
end
|
251
|
+
|
252
|
+
describe '#size' do
|
253
|
+
context 'when initialized by "chr1:400-410"' do
|
254
|
+
it 'returens 11' do
|
255
|
+
Bio::GenomicInterval.parse("chr1:400-410").size.should == 11
|
256
|
+
end
|
257
|
+
end
|
258
|
+
end
|
259
|
+
|
260
|
+
describe '#center' do
|
261
|
+
context 'when initialized by "chr1:11-15"' do
|
262
|
+
it 'returns "chr1:13-13" by the to_s method' do
|
263
|
+
Bio::GenomicInterval.parse("chr1:11-15").center.to_s.should == "chr1:13-13"
|
264
|
+
end
|
265
|
+
end
|
266
|
+
|
267
|
+
context 'when initialize by "chr1:10-15"' do
|
268
|
+
it 'returns "chr1:12-12" by the to_s method' do
|
269
|
+
Bio::GenomicInterval.parse("chr1:10-15").center.to_s.should == "chr1:12-12"
|
270
|
+
end
|
271
|
+
end
|
272
|
+
end
|
273
|
+
|
274
|
+
describe '#overlap for "chr1:400-500"' do
|
275
|
+
context 'given "chr1:100-200"' do
|
276
|
+
it 'returns -199' do
|
277
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-500")
|
278
|
+
subject = Bio::GenomicInterval.parse("chr1:100-200")
|
279
|
+
receiver.overlap(subject).should == -199
|
280
|
+
end
|
281
|
+
end
|
282
|
+
|
283
|
+
context 'given "chr1:100-399"' do
|
284
|
+
it 'returns 0' do
|
285
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-500")
|
286
|
+
subject = Bio::GenomicInterval.parse("chr1:100-399")
|
287
|
+
receiver.overlap(subject).should == 0
|
288
|
+
end
|
289
|
+
end
|
290
|
+
|
291
|
+
context 'given "chr1:100-400"' do
|
292
|
+
it 'returns 1' do
|
293
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-500")
|
294
|
+
subject = Bio::GenomicInterval.parse("chr1:100-400")
|
295
|
+
receiver.overlap(subject).should == 1
|
296
|
+
end
|
297
|
+
end
|
298
|
+
|
299
|
+
context 'given "chr1:410-490"' do
|
300
|
+
it 'returns 81' do
|
301
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-500")
|
302
|
+
subject = Bio::GenomicInterval.parse("chr1:410-490")
|
303
|
+
receiver.overlap(subject).should == 81
|
304
|
+
end
|
305
|
+
end
|
306
|
+
|
307
|
+
context 'given "chr1:300-600"' do
|
308
|
+
it 'returns 101' do
|
309
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-500")
|
310
|
+
subject = Bio::GenomicInterval.parse("chr1:300-600")
|
311
|
+
receiver.overlap(subject).should == 101
|
312
|
+
end
|
313
|
+
end
|
314
|
+
|
315
|
+
context 'given "chr1:400-500"' do
|
316
|
+
it 'returns 101' do
|
317
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-500")
|
318
|
+
subject = Bio::GenomicInterval.parse("chr1:400-500")
|
319
|
+
receiver.overlap(subject).should == 101
|
320
|
+
end
|
321
|
+
end
|
322
|
+
|
323
|
+
context 'given "chr1:450-550"' do
|
324
|
+
it 'returns 51' do
|
325
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-500")
|
326
|
+
subject = Bio::GenomicInterval.parse("chr1:450-550")
|
327
|
+
receiver.overlap(subject).should == 51
|
328
|
+
end
|
329
|
+
end
|
330
|
+
|
331
|
+
context 'given "chr1:501-600"' do
|
332
|
+
it 'returns 0' do
|
333
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-500")
|
334
|
+
subject = Bio::GenomicInterval.parse("chr1:501-600")
|
335
|
+
receiver.overlap(subject).should == 0
|
336
|
+
end
|
337
|
+
end
|
338
|
+
|
339
|
+
context 'given "chr1:550-600"' do
|
340
|
+
it 'returns -49' do
|
341
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-500")
|
342
|
+
subject = Bio::GenomicInterval.parse("chr1:550-600")
|
343
|
+
receiver.overlap(subject).should == -49
|
344
|
+
end
|
345
|
+
end
|
346
|
+
|
347
|
+
context 'given "chrX:100-900"' do
|
348
|
+
it 'returns 0' do
|
349
|
+
receiver = Bio::GenomicInterval.parse("chr1:400-500")
|
350
|
+
subject = Bio::GenomicInterval.parse("chrX:100-900")
|
351
|
+
receiver.overlap(subject).should == 0
|
352
|
+
end
|
353
|
+
end
|
354
|
+
|
355
|
+
end
|
356
|
+
|
357
|
+
end
|
data/spec/spec_helper.rb
ADDED
@@ -0,0 +1,12 @@
|
|
1
|
+
$LOAD_PATH.unshift(File.join(File.dirname(__FILE__), '..', 'lib'))
|
2
|
+
$LOAD_PATH.unshift(File.dirname(__FILE__))
|
3
|
+
require 'rspec'
|
4
|
+
require 'bio-genomic-interval'
|
5
|
+
|
6
|
+
# Requires supporting files with custom matchers and macros, etc,
|
7
|
+
# in ./support/ and its subdirectories.
|
8
|
+
Dir["#{File.dirname(__FILE__)}/support/**/*.rb"].each {|f| require f}
|
9
|
+
|
10
|
+
RSpec.configure do |config|
|
11
|
+
|
12
|
+
end
|
metadata
ADDED
@@ -0,0 +1,157 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: bio-genomic-interval
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
hash: 27
|
5
|
+
prerelease:
|
6
|
+
segments:
|
7
|
+
- 0
|
8
|
+
- 1
|
9
|
+
- 0
|
10
|
+
version: 0.1.0
|
11
|
+
platform: ruby
|
12
|
+
authors:
|
13
|
+
- Hiroyuki Mishima
|
14
|
+
autorequire:
|
15
|
+
bindir: bin
|
16
|
+
cert_chain: []
|
17
|
+
|
18
|
+
date: 2011-04-07 00:00:00 +09:00
|
19
|
+
default_executable:
|
20
|
+
dependencies:
|
21
|
+
- !ruby/object:Gem::Dependency
|
22
|
+
name: rspec
|
23
|
+
requirement: &id001 !ruby/object:Gem::Requirement
|
24
|
+
none: false
|
25
|
+
requirements:
|
26
|
+
- - ~>
|
27
|
+
- !ruby/object:Gem::Version
|
28
|
+
hash: 3
|
29
|
+
segments:
|
30
|
+
- 2
|
31
|
+
- 3
|
32
|
+
- 0
|
33
|
+
version: 2.3.0
|
34
|
+
type: :development
|
35
|
+
prerelease: false
|
36
|
+
version_requirements: *id001
|
37
|
+
- !ruby/object:Gem::Dependency
|
38
|
+
name: bundler
|
39
|
+
requirement: &id002 !ruby/object:Gem::Requirement
|
40
|
+
none: false
|
41
|
+
requirements:
|
42
|
+
- - ~>
|
43
|
+
- !ruby/object:Gem::Version
|
44
|
+
hash: 23
|
45
|
+
segments:
|
46
|
+
- 1
|
47
|
+
- 0
|
48
|
+
- 0
|
49
|
+
version: 1.0.0
|
50
|
+
type: :development
|
51
|
+
prerelease: false
|
52
|
+
version_requirements: *id002
|
53
|
+
- !ruby/object:Gem::Dependency
|
54
|
+
name: jeweler
|
55
|
+
requirement: &id003 !ruby/object:Gem::Requirement
|
56
|
+
none: false
|
57
|
+
requirements:
|
58
|
+
- - ~>
|
59
|
+
- !ruby/object:Gem::Version
|
60
|
+
hash: 7
|
61
|
+
segments:
|
62
|
+
- 1
|
63
|
+
- 5
|
64
|
+
- 2
|
65
|
+
version: 1.5.2
|
66
|
+
type: :development
|
67
|
+
prerelease: false
|
68
|
+
version_requirements: *id003
|
69
|
+
- !ruby/object:Gem::Dependency
|
70
|
+
name: rcov
|
71
|
+
requirement: &id004 !ruby/object:Gem::Requirement
|
72
|
+
none: false
|
73
|
+
requirements:
|
74
|
+
- - ">="
|
75
|
+
- !ruby/object:Gem::Version
|
76
|
+
hash: 3
|
77
|
+
segments:
|
78
|
+
- 0
|
79
|
+
version: "0"
|
80
|
+
type: :development
|
81
|
+
prerelease: false
|
82
|
+
version_requirements: *id004
|
83
|
+
- !ruby/object:Gem::Dependency
|
84
|
+
name: bio
|
85
|
+
requirement: &id005 !ruby/object:Gem::Requirement
|
86
|
+
none: false
|
87
|
+
requirements:
|
88
|
+
- - ">="
|
89
|
+
- !ruby/object:Gem::Version
|
90
|
+
hash: 5
|
91
|
+
segments:
|
92
|
+
- 1
|
93
|
+
- 4
|
94
|
+
- 1
|
95
|
+
version: 1.4.1
|
96
|
+
type: :development
|
97
|
+
prerelease: false
|
98
|
+
version_requirements: *id005
|
99
|
+
description: "a BioRuby plugin: handling genomic intervals,such as \"chr1:123-456\", and overlaps between two intervalsrake "
|
100
|
+
email: missy@be.to
|
101
|
+
executables: []
|
102
|
+
|
103
|
+
extensions: []
|
104
|
+
|
105
|
+
extra_rdoc_files:
|
106
|
+
- LICENSE.txt
|
107
|
+
- README.rdoc
|
108
|
+
files:
|
109
|
+
- .document
|
110
|
+
- .rspec
|
111
|
+
- Gemfile
|
112
|
+
- Gemfile.lock
|
113
|
+
- LICENSE.txt
|
114
|
+
- README.rdoc
|
115
|
+
- Rakefile
|
116
|
+
- VERSION
|
117
|
+
- bio-genomic-interval.gemspec
|
118
|
+
- lib/bio-genomic-interval.rb
|
119
|
+
- spec/bio-genomic-interval_spec.rb
|
120
|
+
- spec/spec_helper.rb
|
121
|
+
has_rdoc: true
|
122
|
+
homepage: http://github.com/misshie/bioruby-genomic-interval
|
123
|
+
licenses:
|
124
|
+
- MIT
|
125
|
+
post_install_message:
|
126
|
+
rdoc_options: []
|
127
|
+
|
128
|
+
require_paths:
|
129
|
+
- lib
|
130
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
131
|
+
none: false
|
132
|
+
requirements:
|
133
|
+
- - ">="
|
134
|
+
- !ruby/object:Gem::Version
|
135
|
+
hash: 3
|
136
|
+
segments:
|
137
|
+
- 0
|
138
|
+
version: "0"
|
139
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
140
|
+
none: false
|
141
|
+
requirements:
|
142
|
+
- - ">="
|
143
|
+
- !ruby/object:Gem::Version
|
144
|
+
hash: 3
|
145
|
+
segments:
|
146
|
+
- 0
|
147
|
+
version: "0"
|
148
|
+
requirements: []
|
149
|
+
|
150
|
+
rubyforge_project:
|
151
|
+
rubygems_version: 1.6.2
|
152
|
+
signing_key:
|
153
|
+
specification_version: 3
|
154
|
+
summary: "a BioRuby plugin: handling genomic interavals and overlaps"
|
155
|
+
test_files:
|
156
|
+
- spec/bio-genomic-interval_spec.rb
|
157
|
+
- spec/spec_helper.rb
|