miga-base 1.3.12.2 → 1.3.13.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: cf660d951441269a671b7fb57d3e91a572fdb62d5ffbaeda9ff412042798759b
-  data.tar.gz: 1721194dcce4a70e4cad66ca4cea5db88f521323022b62c2c93740458a1b0bea
+  metadata.gz: 8659bae6ee7d40a76a464840a56b4aff77f9a848d777349e98d8c316717b0021
+  data.tar.gz: dab1cbd8ee24503b4dfbe0e8b034da195120251c258c798bbba92668a52d3cfa
 SHA512:
-  metadata.gz: ff0b9c3a8db37fa9b0d75a0b884154b416669e068eab9f22f406facf9822345a38baccc224138a15fb5e9e1dff786e55e2ebc0d870573b892f056bf7e18e275a
-  data.tar.gz: af7dcea647e3ead8fdbde6f917264efd93335cc9c7878bd7a4b43b8dbb27434882dff76728fa984c5724802a581421ddd333f6cfe46ab1d6c75fb4bb953bc378
+  metadata.gz: '09a916e1db7bb4912111e83c2d832a779a245c8eb326aee3cfcead8611d484f737a1cbcae1a1279a2673a2f31a8252c859b533b7ae9b6c74c8881b41e63e7c40'
+  data.tar.gz: 43e7065632d2642abb0c6bb5fac8ae96e0696c0f48b128376fa5d1ac929c1c7b787df61af78bee5a4824049da7b2baf976f9bb3d3f083184ca02deefc166114e

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

@@ -5,6 +5,9 @@ require 'miga/cli/action'
 require 'miga/remote_dataset'
 
 class MiGA::Cli::Action::Get < MiGA::Cli::Action
+  require 'miga/cli/action/download/base'
+  include MiGA::Cli::Action::Download::Base
+
   def parse_cli
     cli.defaults = {
       query: false, universe: :ncbi, db: :nuccore, get_md: false, only_md: false
@@ -69,12 +72,18 @@ class MiGA::Cli::Action::Get < MiGA::Cli::Action
         '--api-key STRING',
         'API key for the given universe'
       ) { |v| cli[:api_key] = v }
+      opt.on(
+        '--ncbi-taxonomy-dump [path]',
+        'Path to an NCBI Taxonomy dump directory to query instead of API calls',
+        'If the path is not passed, the dump is automatically downloaded'
+      ) { |v| cli[:ncbi_taxonomy_dump] = v || true }
     end
   end
 
   def perform
     glob = get_sub_cli
     p = cli.load_project
+    load_ncbi_taxonomy_dump
     glob.each do |sub_cli|
       rd = create_remote_dataset(sub_cli, p)
       next if rd.nil?

data/lib/miga/remote_dataset/download.rb

@@ -96,9 +96,21 @@ class MiGA::RemoteDataset
       o = download_rest(opts.merge(universe: :ncbi, db: :nuccore))
       return o unless o.strip.empty?
 
-      MiGA::MiGA.DEBUG 'Empty sequence, attempting download from NCBI assembly'
-      opts[:format] = :fasta
-      ncbi_asm_get(opts)
+      begin
+        MiGA::MiGA.DEBUG 'Empty sequence, attempting download as NCBI assembly'
+        opts[:format] = :fasta
+        ncbi_asm_get(opts)
+      rescue => e
+        raise e unless opts[:obj]&.metadata&.dig(:ncbi_wgs)
+        MiGA::MiGA.DEBUG e.to_s
+      end
+
+      MiGA::MiGA.DEBUG 'Empty sequence, attempting download as WGS records'
+      a, b = opts[:obj].metadata[:ncbi_wgs].split('-', 2)
+      pref = longest_common_prefix([a, b])
+      rang = a[pref.size .. -1].to_i .. b[pref.size .. -1].to_i
+      ids  = rang.map { |k| "%s%0#{a.size - pref.size}i" % [pref, k] }
+      download_rest(opts.merge(universe: :ncbi, db: :nuccore, ids: ids))
     end
 
     ##
@@ -184,6 +196,15 @@ class MiGA::RemoteDataset
       end
       tree
     end
+
+    ##
+    # From: https://github.com/isisAnchalee/Algorithms
+    def longest_common_prefix(strs)
+      return '' if strs.empty?
+      min, max = strs.minmax
+      idx = min.size.times { |i| break i if min[i] != max[i] }
+      min[0...idx]
+    end
   end
 end
 

data/lib/miga/remote_dataset.rb

@@ -322,9 +322,16 @@ class MiGA::RemoteDataset < MiGA::MiGA
     taxid = metadata.dig(:ncbi_dataset, :organism, :tax_id)
     return taxid if taxid
 
-    # Try from GenBank document (obtain it)
+    # Get GenBank document
     doc = self.class.download(:ncbi, db, ids, :gb, nil, {}, self).split(/\n/)
-    ln = doc.grep(%r{^\s+/db_xref="taxon:}).first
+
+    # Since we're here, try to recover WGS for synthetic records
+    ln  = doc.grep(/^WGS\s+\S+-\S+/).first
+    wgs = ln&.gsub(/^WGS\s+(\S+-\S+).*/, '\1')
+    @metadata[:ncbi_wgs] = wgs if wgs
+
+    # Now try to extract taxid from GenBank
+    ln  = doc.grep(%r{^\s+/db_xref="taxon:}).first
     return nil if ln.nil?
 
     ln.sub!(/.*(?:"taxon:)(\d+)["; ].*/, '\\1')

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, 12, 2].freeze
+  VERSION = [1.3, 13, 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, 3, 8)
+  VERSION_DATE = Date.new(2024, 3, 10)
 
   ##
   # References of MiGA

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: miga-base
 version: !ruby/object:Gem::Version
-  version: 1.3.12.2
+  version: 1.3.13.0
 platform: ruby
 authors:
 - Luis M. Rodriguez-R
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2024-03-08 00:00:00.000000000 Z
+date: 2024-03-10 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: daemons