Merge branch 'postreleasefix/110' into release/110

Ensembl · May 18, 2023 · 498c619 · 498c619
2 parents 8bfc5fb + abcb2b3
commit 498c619
Show file tree

Hide file tree

Showing 4 changed files with 27 additions and 15 deletions.
diff --git a/modules/Bio/EnsEMBL/Variation/DBSQL/VariationFeatureAdaptor.pm b/modules/Bio/EnsEMBL/Variation/DBSQL/VariationFeatureAdaptor.pm
@@ -2068,10 +2068,8 @@ sub _hgvs_from_components {
 
   # take alternate allele from genomic reference & coordinates if not supplied in HGVS string for a duplication
   if($description =~ /dup/){ 
-     ## special case: handle as insertion for ensembl object purposes 
-     $start = $end ;
-     if($strand  == 1){ $start++; }
-     else{             $end--;   }
+    ## special case: handle as insertion for ensembl object purposes 
+    $start = $end + 1; 
 
     $ref_allele = "-" ;
     $alt_allele = $refseq_allele;

diff --git a/modules/Bio/EnsEMBL/Variation/TranscriptVariationAllele.pm b/modules/Bio/EnsEMBL/Variation/TranscriptVariationAllele.pm
@@ -1387,14 +1387,17 @@ sub hgvs_transcript {
 
   return undef if (($self->{_slice}->end - $self->{_slice}->start + 1) < ($self->{_slice_end} + $offset_to_add));
   #return undef if (length($self->{_slice}->seq()) < ($self->{_slice_end} + $offset_to_add));
+
+  my $dup_lookup_direction = ($refseq_strand == -1 && !$offset_to_add) ? 1 : -1;
   $hgvs_notation = hgvs_variant_notation(
     $variation_feature_sequence,    ### alt_allele,
     $self->{_slice}->seq(),                             ### using this to extract ref allele
     $self->{_slice_start} + $offset_to_add,
     $self->{_slice_end} + $offset_to_add,
     "",
     "",
-    $var_name 
+    $var_name,
+    $dup_lookup_direction
   );
 
   ### This should not happen
@@ -1403,6 +1406,7 @@ sub hgvs_transcript {
     return undef;
   } 
 
+
   ## check for the same bases in ref and alt strings before or after the variant
   $hgvs_notation = _clip_alleles($hgvs_notation) unless $hgvs_notation->{'type'} eq 'dup';
 

diff --git a/modules/Bio/EnsEMBL/Variation/Utils/Sequence.pm b/modules/Bio/EnsEMBL/Variation/Utils/Sequence.pm
@@ -498,11 +498,15 @@ sub hgvs_variant_notation {
   my $display_start = shift;
   my $display_end = shift;
   my $var_name  = shift;
+  my $dup_lookup_direction = shift;
 
   # If display_start and display_end were not specified, use ref_start and ref_end
   $display_start ||= $ref_start;
   $display_end ||= $ref_end;
 
+  # the direction to look for ref seq match with alt allele to see if it is dup type
+  $dup_lookup_direction ||= -1;
+
   #Throw an exception if the lengths of the display interval and reference interval are different
   throw("The coordinate interval for display is of different length than for the reference allele") if (($display_end - $display_start) != ($ref_end - $ref_start));
 
@@ -526,7 +530,6 @@ sub hgvs_variant_notation {
     #warn "\nError in HGVS calculation for $var_name: alt allele ($alt_allele) is the same as the reference allele ($ref_allele) - potential strand or allele ordering problem - skipping\n";
     return undef ;
   }
-
   # Store the notation in a hash that will be returned
   my %notation;
   $notation{'start'} = $display_start;
@@ -567,14 +570,20 @@ sub hgvs_variant_notation {
   if (!$ref_length) {
 
     # Get the same number of nucleotides preceding the insertion as the length of the insertion
-    my $prev_str = substr($ref_sequence,($ref_end-$alt_length),$alt_length);
+    my $prev_str = $dup_lookup_direction == -1 ? 
+      substr($ref_sequence,($ref_end-$alt_length),$alt_length) :
+      substr($ref_sequence,$ref_end,$alt_length);
 
     # If they match, this is a duplication
     if ($prev_str eq $alt_allele) {
-
-      $notation{'start'} = ($display_end - $alt_length + 1);
+
+      if ($dup_lookup_direction == -1){
+        $notation{'start'} = ($display_end - $alt_length + 1);
+      }
+      else{
+        $notation{'end'} = $display_start + $alt_length - 1;
+      }
       $notation{'type'} = 'dup';
-
       # Return the notation
       return \%notation;
     }
@@ -605,6 +614,7 @@ sub hgvs_variant_notation {
 
   # Else, it's gotta be a delins
   $notation{'type'} = 'delins';
+
   return \%notation;
 }
 

diff --git a/modules/t/hgvs_parser.t b/modules/t/hgvs_parser.t
@@ -486,7 +486,7 @@ my %test_input = (
 );
 
 my %test_input_shifted = (
-     28 => ["X:g.131215392_131215393insA",
+     28 => ["X:g.131215393_131215394insA",
             "ENST00000298542.4:c.905+998dup", 
             "X:g.131215401dup",
            ],    
@@ -530,9 +530,9 @@ my %test_input_shifted = (
 
 ## results which change on left-shifting - not shifted
 my %test_output_no_shift = ( 
-     1 => ["X:g.131215393dup",
+     1 => ["X:g.131215394dup",
            "A",
-           "ENST00000298542.4:c.905+997dup", 
+           "ENST00000298542.4:c.905+997dup",
            "T",
            "",
            "duplication, intronic rc transcript"
@@ -586,8 +586,8 @@ my %test_output_no_shift = (
 
 
 my %test_input_no_shift = ( 
-     1 => ["X:g.131215393dup",
-           "ENST00000298542.4:c.905+997dup", 
+     1 => ["X:g.131215394dup",
+           "ENST00000298542.4:c.905+997dup",
           ],    
      2 => ["NC_000011.9:g.32417913_32417914insCCTACGAGTACTACC",
            "ENST00000530998.1:c.451_452insGGTAGTACTCGTAGG",