spec/cli_spec.rb in lederhosen-1.8.2 vs spec/cli_spec.rb in lederhosen-2.0.0

@@ -23,37 +23,89 @@
     $?.success?.should be_true
     File.exists?(File.join($test_dir, 'clusters.uc')).should be_false
   end
 
   it 'can cluster reads using usearch' do
-    `./bin/lederhosen cluster --input spec/data/trimmed/ILT_L_9_B_001.fasta --database #{$test_dir}/test_db.udb --identity 0.95 --output #{$test_dir}/clusters.uc`
+    `./bin/lederhosen cluster --input spec/data/trimmed/ILT_L_9_B_001.fasta --database #{$test_dir}/test_db.udb --identity 0.99 --output #{$test_dir}/clusters.uc`
     $?.success?.should be_true
     File.exists?(File.join($test_dir, 'clusters.uc')).should be_true
   end
 
-  it 'should build abundance matrices for each level' do
-    levels = "domain phylum class order FAMILY genus Species"
-    `./bin/lederhosen otu_table --files=spec/data/test.uc --prefix=#{$test_dir}/otu_table --levels=#{levels}`
+  it 'can separate unclassified reads from usearch output' do
+    `./bin/lederhosen separate_unclassified --uc-file=spec/data/test.uc --reads=spec/data/trimmed/ILT_L_9_B_001.fasta --output=#{$test_dir}/unclassified.fasta`
     $?.success?.should be_true
+    unclassified_results = File.readlines("spec/data/test.uc")
+                               .select { |x| x =~ /^N/ }
+                               .size
+    unclassified_reads = File.readlines("#{$test_dir}/unclassified.fasta")
+                             .select { |x| x =~ /^>/ }
+                             .size
+  
+    unclassified_results.should == unclassified_reads
   end
+  
+  it 'can separate unclassified reads from usearch output using strict pairing' do
+    `./bin/lederhosen separate_unclassified --strict=genus --uc-file=spec/data/test.uc --reads=spec/data/trimmed/ILT_L_9_B_001.fasta --output=#{$test_dir}/unclassified.strict_genus.fasta`
+    $?.success?.should be_true
+    File.readlines("#{$test_dir}/unclassified.strict_genus.fasta")
+      .select { |x| x =~ /^>/ }
+      .size.should be_even
+  end
 
-  it 'should filter OTU abundance matrices' do
-    `./bin/lederhosen otu_filter --input=#{$test_dir}/otu_table.species.csv --output=#{$test_dir}/otu_table.filtered.csv --reads 1 --samples 1`
+  it 'can create taxonomy count tables' do
+    `./bin/lederhosen count_taxonomies --input=spec/data/test.uc --output=#{$test_dir}/taxonomy_count.txt`
     $?.success?.should be_true
+    File.exists?(File.join($test_dir, 'taxonomy_count.txt')).should be_true
   end
 
-  it 'should combine OTU abundance matrices' do
-    `./bin/lederhosen join_otu_tables --input=#{$test_dir}/otu_table*.csv --output=#{$test_dir}/merged.csv`
+  it 'generates taxonomy tables w/ comma-free taxonomic descriptions' do
+    File.readlines(File.join($test_dir, 'taxonomy_count.txt'))
+      .map(&:strip)
+      .map { |x| x.count(',') }
+      .uniq
+      .should == [1]
+  end
+
+  %w{domain phylum class order family genus species}.each do |level|
+    it "generates taxonomy tables only counting pairs that agree at level: #{level}" do
+      `./bin/lederhosen count_taxonomies --input=spec/data/test.uc --output=#{$test_dir}/taxonomy_count.strict.#{level}.txt --strict=#{level}`
+      $?.success?.should be_true
+
+      lines = File.readlines(File.join($test_dir, "taxonomy_count.strict.#{level}.txt"))
+
+      # make sure total number of reads is even
+      # requires that there should be an odd number if classification is not strict
+      lines.select { |x| !(x =~ /^#/) }
+           .map(&:strip)
+           .map { |x| x.split(',') }
+           .map(&:last)
+           .map(&:to_i)
+           .inject(:+).should be_even
+    end
+  end
+
+  %w{domain phylum class order family genus species}.each do |level|
+    it "should create OTU abundance matrices from taxonomy count tables at level: #{level}" do
+      `./bin/lederhosen otu_table --files=#{$test_dir}/taxonomy_count.strict.*.txt --level=#{level} --output=#{$test_dir}/otus_genus.strict.csv`
+      $?.success?.should be_true
+    end
+  end
+
+  it 'should filter OTU abundance matrices' do
+    # TODO
+    # filtering should move filtered reads to 'unclassified_reads' so that we maintain
+    # our knowledge of depth of coverage throughout
+    # this makes normalization better later.
+    `./bin/lederhosen otu_filter --input=#{$test_dir}/otus_genus.strict.csv --output=#{$test_dir}/otu_table.filtered.csv --reads 1 --samples 1`
     $?.success?.should be_true
   end
 
   it 'should split a fasta file into smaller fasta files (optionally gzipped)' do
     `./bin/lederhosen split_fasta --input=spec/data/trimmed/ILT_L_9_B_001.fasta --out-dir=#{$test_dir}/split/ --gzip true -n 100`
     $?.success?.should be_true
   end
 
   it 'should print representative sequences from uc files' do
     `./bin/lederhosen get_reps --input=#{$test_dir}/clusters.uc --database=spec/data/trimmed/ILT_L_9_B_001.fasta --output=#{$test_dir}/representatives.fasta`
+    $?.success?.should be_true
   end
-
-  it 'should create a fasta file containing representative reads for each cluster'
 end