From 3eda626bd613909e151b9b69c0221945bbd9e475 Mon Sep 17 00:00:00 2001
From: Nuno Agostinho <nuno@ebi.ac.uk>
Date: Thu, 14 Mar 2024 13:25:38 +0000
Subject: [PATCH 1/3] Fix SVTYPE when using IUPAC nucleotide codes

---
 modules/Bio/EnsEMBL/VEP/Parser/VCF.pm | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/modules/Bio/EnsEMBL/VEP/Parser/VCF.pm b/modules/Bio/EnsEMBL/VEP/Parser/VCF.pm
index 8cebd55e8..bb5972e12 100644
--- a/modules/Bio/EnsEMBL/VEP/Parser/VCF.pm
+++ b/modules/Bio/EnsEMBL/VEP/Parser/VCF.pm
@@ -460,9 +460,15 @@ sub create_StructuralVariationFeatures {
     $parser->get_IDs,
   );
 
-  ## get structural variant type from SVTYPE tag (deprecated in VCF 4.4) or ALT
+  ## get structural variant type from ALT or (deprecated) SVTYPE tag
   my $alt = join("/", @$alts);
-  my $type = $alt ne '.' ? $alt : $info->{SVTYPE};
+  my $type = $alt;
+
+  # replace with SVTYPE tag if ALT does not follow VCF 4.4 specs
+  if ($info->{SVTYPE} && $alt !~ /^<?(DEL|INS|DUP|INV|CNV|CN=?[0-9]+)/) {
+    $type = $info->{SVTYPE};
+  }
+
   my $so_term = $self->get_SO_term($type);
   unless ($so_term) {
     $skip_line = 1;

From 0b706f2e21c46b22c70447961da88ff62961e2bb Mon Sep 17 00:00:00 2001
From: Nuno Agostinho <nuno@ebi.ac.uk>
Date: Thu, 14 Mar 2024 13:38:08 +0000
Subject: [PATCH 2/3] Improve warning message for unsupported SV types

---
 modules/Bio/EnsEMBL/VEP/Parser.pm | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/modules/Bio/EnsEMBL/VEP/Parser.pm b/modules/Bio/EnsEMBL/VEP/Parser.pm
index b328db8d5..5270770f4 100755
--- a/modules/Bio/EnsEMBL/VEP/Parser.pm
+++ b/modules/Bio/EnsEMBL/VEP/Parser.pm
@@ -723,7 +723,7 @@ sub get_SO_term {
   my $res = $terms{$abbrev};
   ## unsupported SV types
   if ($self->isa('Bio::EnsEMBL::VEP::Parser')) {
-    $self->skipped_variant_msg("$abbrev type is not supported") unless $res;
+    $self->skipped_variant_msg("$abbrev is not a supported structural variant type") unless $res;
   }
   return $res;
 }

From 486f60e56b70c0a76a5f7157876c4a0625953b1d Mon Sep 17 00:00:00 2001
From: Nuno Agostinho <nuno@ebi.ac.uk>
Date: Fri, 15 Mar 2024 11:45:25 +0000
Subject: [PATCH 3/3] Fix unit tests

---
 t/Parser_VCF.t | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/t/Parser_VCF.t b/t/Parser_VCF.t
index 8c85691f8..86f800f20 100755
--- a/t/Parser_VCF.t
+++ b/t/Parser_VCF.t
@@ -621,14 +621,14 @@ is_deeply($cvf, bless( {
                  'inner_start' => '774570',
                  'strand' => 1,
                  'seq_region_end' => 828435,
-                 'class_SO_term' => '<CPX>',
+                 'class_SO_term' => 'CPX',
                  'variation_name' => 'gnomAD_v2_CPX_1_1',
                  'start' => 774570
                },
                 'Bio::EnsEMBL::Variation::StructuralVariationFeature' ) , 'StructuralVariationFeature - CPX skipped');
 
 
-like($tmp, qr/CPX type is not supported/, 'StructuralVariationFeature - skip CPX warning');
+like($tmp, qr/CPX is not a supported structural variant type/, 'StructuralVariationFeature - skip CPX warning');
 
 open(STDERR, ">&SAVE") or die "Can't restore STDERR\n";