miga-base 1.3.18.0 → 1.3.20.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: bf9ebc2ca06846d9b76c981c2a89896080690bc44d979ba28b4a9173f7386c7a
-  data.tar.gz: a156a4f0894f1988862b8daf296b916c80cf197eb36fb7c13c4698975403c35e
+  metadata.gz: c63c60b5324c31420fa616747435b2e33e6fc7d0855cab55381f139789ccedc1
+  data.tar.gz: 5cbbd4de1f66b452ea72e4dface8617606aec919af08aaddea1a0f05e745873a
 SHA512:
-  metadata.gz: b6e1318553d24d93af467670754e86e5281c7418d4242ad11b56125889188fb7cde48a1858f20e1b07bab04a372c4aac718e04c005bdfadccf641c15cd7cf2ed
-  data.tar.gz: e584f1b4442a099fb16c4b0c07515b9332867c5b72438bae9e33467f9d0911c7bd4811a6861121482cada1b88f44aaec2efd6d2f668fa6a05c722c9fe56995bc
+  metadata.gz: 57c3abcf3030051fb6d9839b58f0f9bfecd75aff18d8deb1f0b9d7d5244d6103b9764de844409d5dbec2ffffcffadaeff0b4a804419d21e8a74e09caf5ee7f73
+  data.tar.gz: bef9383f052d1331b0154b61d7c9e8dba07194775e5d05aea110771c26597fd6549ed4884d9a4c92c8484952c8f26d33f924281f56957e01d3110cb81c3d2d98

data/lib/miga/cli/action/relatives.rb

@@ -0,0 +1,45 @@
+# @package MiGA
+# @license Artistic-2.0
+
+require 'miga/cli/action'
+
+class MiGA::Cli::Action::Relatives < MiGA::Cli::Action
+  def parse_cli
+    cli.defaults = { metric: :aai, external: false, how_many: 5 }
+    cli.parse do |opt|
+      cli.opt_object(opt, [:project, :dataset_opt])
+      cli.opt_filter_datasets(opt)
+      opt.on(
+        '--metric STRING',
+        'Use this metric of identity, one of ANI or AAI (default)'
+      ) { |v| cli[:metric] = v.downcase.to_sym }
+      opt.on(
+        '--ref-project',
+        'Report distances to the external reference project used for taxonomy',
+        'By default: report distances to other datasets in the project'
+      ) { |v| cli[:external] = v }
+      opt.on(
+        '-n', '--how-many INT', Integer,
+        'Number of top values to report'
+      ) { |v| cli[:how_many] = v }
+      opt.on(
+        '--tab',
+        'Return a tab-delimited table'
+      ) { |v| cli[:tabular] = v }
+      opt.on(
+        '-o', '--output PATH',
+        'Create output file instead of returning to STDOUT'
+      ) { |v| cli[:output] = v }
+    end
+  end
+
+  def perform
+    cr = []
+    cli.load_and_filter_datasets.each do |d|
+      d_cr = d.closest_relatives(cli[:how_many], cli[:external], cli[:metric])
+      cr += d_cr.map { |i| [d.name] + i }
+    end
+    io = cli[:output].nil? ? $stdout : File.open(cli[:output], 'w')
+    cli.table(['dataset_A', 'dataset_B', cli[:metric]], cr, io)
+  end
+end

data/lib/miga/common/format.rb

@@ -23,8 +23,9 @@ module MiGA::Common::Format
   end
 
   ##
-  # Cleans a FastA file in place.
-  def clean_fasta_file(file)
+  # Cleans a FastA file in place, removing all sequences shorter than
+  # +min_len+
+  def clean_fasta_file(file, min_len = 1)
     tmp_fh = nil
     tmp_path = nil
     begin
@@ -39,19 +40,24 @@ module MiGA::Common::Format
         tmp_path = tmp_fh.path
         fh = File.open(file, 'r')
       end
-      buffer = ''.dup
+      next_seq = ['', '']
       fh.each_line do |ln|
         ln.chomp!
         if ln =~ /^>\s*(\S+)(.*)/
           id, df = $1, $2
-          tmp_fh.print buffer.wrap_width(80)
-          buffer = ''.dup
-          tmp_fh.puts ">#{id.gsub(/[^A-Za-z0-9_\|\.]/, '_')}#{df}"
+          if next_seq[1].length >= min_len
+            tmp_fh.puts next_seq[0]
+            tmp_fh.print next_seq[1].wrap_width(80)
+          end
+          next_seq = [">#{id.gsub(/[^A-Za-z0-9_\|\.]/, '_')}#{df}", '']
         else
-          buffer << ln.gsub(/[^A-Za-z\.\-]/, '')
+          next_seq[1] += ln.gsub(/[^A-Za-z\.\-]/, '')
         end
       end
-      tmp_fh.print buffer.wrap_width(80)
+      if next_seq[1].length >= min_len
+        tmp_fh.puts next_seq[0]
+        tmp_fh.print next_seq[1].wrap_width(80)
+      end
       tmp_fh.close
       fh.close
       FileUtils.mv(tmp_path, file)

data/lib/miga/dataset/result/add.rb

@@ -84,7 +84,7 @@ module MiGA::Dataset::Result::Add
     opts[:is_clean] ||= false
     r.clean! if opts[:is_clean]
     unless r.clean?
-      MiGA::MiGA.clean_fasta_file(r.file_path(:largecontigs))
+      MiGA::MiGA.clean_fasta_file(r.file_path(:largecontigs), 1000)
       r.clean!
     end
     r

data/lib/miga/dataset.rb

@@ -177,22 +177,23 @@ class MiGA::Dataset < MiGA::MiGA
   alias is_active? active?
 
   ##
-  # Returns an Array of +how_many+ duples (Arrays) sorted by AAI:
-  # - +0+: A String with the name(s) of the reference dataset.
-  # - +1+: A Float with the AAI.
-  # This function is currently only supported for query datasets when
-  # +ref_project+ is false (default), and only for reference dataset when
-  # +ref_project+ is true. It returns +nil+ if this analysis is not supported.
-  def closest_relatives(how_many = 1, ref_project = false)
-    return nil if (ref? != ref_project) || multi?
+  # Returns an Array of +how_many+ duples (Arrays) sorted by +metric+
+  # (one of +:aai+ or +:ani+):
+  # - +0+: A String with the name(s) of the reference dataset
+  # - +1+: A Float with the AAI/ANI
+  def closest_relatives(how_many = 1, ref_project = false, metric = :aai)
+    return nil if multi?
 
     r = result(ref_project ? :taxonomy : :distances)
     return nil if r.nil?
 
     require 'miga/sqlite'
-    MiGA::SQLite.new(r.file_path(:aai_db)).run(
-      'SELECT seq2, aai FROM aai WHERE seq2 != ? ' \
-      'GROUP BY seq2 ORDER BY aai DESC LIMIT ?', [name, how_many]
+    metric = metric.to_s.downcase.to_sym
+    db_key = :"#{metric}_db"
+    MiGA::SQLite.new(r.file_path(db_key)).run(
+      "SELECT seq2, #{metric} FROM #{metric} WHERE seq2 != ? " \
+      "GROUP BY seq2 ORDER BY #{metric} DESC LIMIT ?",
+      [name, how_many]
     )
   end
 end

data/lib/miga/version.rb

@@ -12,7 +12,7 @@ module MiGA
   # - String indicating release status:
   #   - rc* release candidate, not released as gem
   #   - [0-9]+ stable release, released as gem
-  VERSION = [1.3, 18, 0].freeze
+  VERSION = [1.3, 20, 0].freeze
 
   ##
   # Nickname for the current major.minor version.
@@ -20,7 +20,7 @@ module MiGA
 
   ##
   # Date of the current gem relese.
-  VERSION_DATE = Date.new(2024, 7, 12)
+  VERSION_DATE = Date.new(2024, 7, 23)
 
   ##
   # References of MiGA

data/scripts/assembly.bash

@@ -58,7 +58,7 @@ if [[ -s "$DATASET/scaffold.fa" ]] ; then
 else
   ln -s "$DATASET/contig.fa" "$DATASET.AllContigs.fna"
 fi
-FastA.length.pl "$DATASET.AllContigs.fna" | awk '$2>=1000{print $1}' \
+FastA.length.pl "$DATASET.AllContigs.fna" | awk '$2 >= 1000 { print $1 }' \
   | FastA.filter.pl /dev/stdin "$DATASET.AllContigs.fna" \
   > "$DATASET.LargeContigs.fna"
 

data/test/result_stats_test.rb

@@ -83,7 +83,7 @@ class ResultStatsTest < Test::Unit::TestCase
     # Prepare result
     dir = 'data/05.assembly'
     fa = file_path(dir, '.LargeContigs.fna')
-    File.open(fa, 'w') { |fh| fh.puts '>1', 'ACTAC' }
+    File.open(fa, 'w') { |fh| fh.puts('>1', 'ACTAC' * 500) }
     touch_done(dir)
     r = dataset.add_result(:assembly)
 
@@ -91,15 +91,29 @@ class ResultStatsTest < Test::Unit::TestCase
     assert_equal({}, r[:stats])
     r.compute_stats
     assert_equal(1, r[:stats][:contigs])
-    assert_equal([5, 'bp'], r[:stats][:total_length])
+    assert_equal([2500, 'bp'], r[:stats][:total_length])
     assert_equal([40.0, '%'], r[:stats][:g_c_content])
   end
 
+  def test_large_contigs
+    # Prepare result
+    dir = 'data/05.assembly'
+    fa = file_path(dir, '.LargeContigs.fna')
+    File.open(fa, 'w') { |fh| fh.puts('>1', 'ACTAC' * 50) }
+    touch_done(dir)
+    r = dataset.add_result(:assembly)
+
+    # Test assertions
+    assert_equal({}, r[:stats])
+    r.compute_stats
+    assert_equal(0, r[:stats][:contigs])
+  end
+
   def test_cds
     # Prepare result
     dir = 'data/06.cds'
     fa = file_path(dir, '.faa')
-    File.open(fa, 'w') { |fh| fh.puts '>1', 'M' }
+    File.open(fa, 'w') { |fh| fh.puts('>1', 'M' * 500) }
     gff = file_path(dir, '.gff3.gz')
     Zlib::GzipWriter.open(gff) do |fh|
       fh.puts '# Model Data: a=b;transl_table=11;'
@@ -111,7 +125,7 @@ class ResultStatsTest < Test::Unit::TestCase
     assert_equal({}, r[:stats])
     r.compute_stats
     assert_equal(1, r[:stats][:predicted_proteins])
-    assert_equal([1.0, 'aa'], r[:stats][:average_length])
+    assert_equal([500.0, 'aa'], r[:stats][:average_length])
     assert_nil(r[:stats][:coding_density])
     test_assembly
     r.compute_stats

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: miga-base
 version: !ruby/object:Gem::Version
-  version: 1.3.18.0
+  version: 1.3.20.0
 platform: ruby
 authors:
 - Luis M. Rodriguez-R
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2024-07-12 00:00:00.000000000 Z
+date: 2024-07-23 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: daemons
@@ -212,6 +212,7 @@ files:
 - lib/miga/cli/action/option.rb
 - lib/miga/cli/action/preproc_wf.rb
 - lib/miga/cli/action/quality_wf.rb
+- lib/miga/cli/action/relatives.rb
 - lib/miga/cli/action/rm.rb
 - lib/miga/cli/action/run.rb
 - lib/miga/cli/action/seqcode_get.rb