Fix check and add tests

Ensembl · Sep 22, 2023 · b552c4c · b552c4c
1 parent 4b70a58
commit b552c4c
Show file tree

Hide file tree

Showing 2 changed files with 50 additions and 8 deletions.
diff --git a/modules/Bio/EnsEMBL/VEP/Parser/VCF.pm b/modules/Bio/EnsEMBL/VEP/Parser/VCF.pm
@@ -609,11 +609,10 @@ sub create_individual_VariationFeatures {
 
   # Compare sample names
   if(lc($self->{individual}->[0]) ne 'all') {
-    my $samples = join(",", sort @{$parser->get_samples});
-    my $input_samples = join(",", sort @{$include});
+    my $found = _find_in_array($parser->get_samples, $include);
 
-    if($samples ne $input_samples) {
-      die("ERROR: Sample IDs given ($input_samples) do not match samples from VCF ($samples)\n");
+    if(!$found) {
+      die("ERROR: Sample IDs given (", join(",", @{$include}), ") do not match samples from VCF (", join(",", @{$parser->get_samples}), ")\n");
     }
   }
 
@@ -698,11 +697,10 @@ sub create_individuals_zyg_VariationFeature {
 
   # Compare sample names
   if(lc($self->{individual_zyg}->[0]) ne 'all') {
-    my $samples = join(",", sort @{$parser->get_samples});
-    my $input_samples = join(",", sort @{$include});
+    my $found = _find_in_array($parser->get_samples, $include);
 
-    if($samples ne $input_samples) {
-      die("ERROR: Sample IDs given ($input_samples) do not match samples from VCF ($samples)\n");
+    if(!$found) {
+      die("ERROR: Sample IDs given (", join(",", @{$include}), ") do not match samples from VCF (", join(",", @{$parser->get_samples}), ")\n");
     }
   }
 
@@ -743,4 +741,20 @@ sub create_individuals_zyg_VariationFeature {
   return \@return;
 }
 
+sub _find_in_array {
+  my $all_samples = shift;
+  my $samples = shift;
+
+  my %all = map { $_ => 1 } @{$all_samples};
+  my %input_samples = map { $_ => 1 } @{$samples};
+
+  foreach my $sample (keys %input_samples) {
+    if(!$all{$sample}) {
+      return 0;
+    }
+  }
+
+  return 1;
+}
+
 1;
diff --git a/t/OutputFactory.t b/t/OutputFactory.t
@@ -2013,6 +2013,34 @@ my $genotype = join ',', sort(@{$result->{ZYG}});
 is($genotype, 'barry:HOM,dave:HET,jeff:HOMREF', 'VariationFeature_to_output_hash - individual_zyg');
 delete($of->{individual_zyg});
 
+### individual_zyg wrong sample name
+dies_ok { get_annotated_buffer({
+  input_file => $test_cfg->create_input_file([
+    ['##fileformat=VCFv4.1'],
+    [qw(#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT dave barry jeff)],
+    [qw(21 25607429 indtest A G . . . GT 0|1 1/1 0/0)],
+  ]),
+  individual_zyg => 'john',
+});
+} 'VariationFeature_to_output_hash - individual_zyg wrong sample name';
+
+### individual_zyg correct sample name
+$ib = get_annotated_buffer({
+  input_file => $test_cfg->create_input_file([
+    ['##fileformat=VCFv4.1'],
+    [qw(#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT dave barry jeff)],
+    [qw(21 25607429 indtest A G . . . GT 0|1 1/1 0/0)],
+  ]),
+  individual_zyg => 'dave,barry',
+});
+
+$of->{individual_zyg} = ['dave,barry'];
+my $result = $of->VariationFeature_to_output_hash($ib->buffer->[0]);
+my $genotype = join ',', sort(@{$result->{ZYG}});
+
+is($genotype, 'barry:HOM,dave:HET', 'VariationFeature_to_output_hash - individual_zyg correct sample name');
+delete($of->{individual_zyg});
+
 
 # done
 done_testing();