Add tweaks to genes2transcripts to handle gene symbols that are updat…

…ing and HGNC genes with no transcript info openvar/rest_variantValidator#186 and also handle the longer deletions in #651

VariantValidator/modules/gapped_mapping.py

@@ -442,7 +442,7 @@ def gapped_g_to_c(self, rel_var, select_transcripts_dict):
                                                                    self.validator.hp,
                                                                    self.validator.vm,
                                                                    self.validator.merge_hgvs_3pr,
-                                                                   genomic_ac=hgvs_genomic_variant.ac,)
+                                                                   genomic_ac=hgvs_genomic_variant.ac)
 
                         if vcf__dict['needs_a_push'] is True:
                             needs_a_push = True
@@ -688,7 +688,8 @@ def gapped_g_to_c(self, rel_var, select_transcripts_dict):
                                 hgvs_not_delins = saved_hgvs_coding
                                 self.disparity_deletion_in = ['false', 'false']
                             elif 'Normalization of intronic variants is not supported' in error:
-                                # We know that this cannot be because of an intronic variant, so must be aligned to tx gap
+                                # We know that this cannot be because of an intronic variant, so must be aligned to
+                                # tx gap
                                 self.disparity_deletion_in = ['transcript', 'Requires Analysis']
                         logger.info(error)
 
@@ -704,7 +705,6 @@ def gapped_g_to_c(self, rel_var, select_transcripts_dict):
 
                     # GAP IN THE TRANSCRIPT DISPARITY DETECTED
                     if self.disparity_deletion_in[0] == 'transcript':
-
                         # Check for issue https://github.com/openvar/variantValidator/issues/385 where the gap is
                         # being identified but oddly the vm is not compensating, likely due to odd sequence
                         try:
@@ -724,7 +724,8 @@ def gapped_g_to_c(self, rel_var, select_transcripts_dict):
                             # the actual length should be == gen_len_difference - 1 not == gen_len_difference
                             if tx_len_difference - self.disparity_deletion_in[1] == gen_len_difference:
                                 # So here we know we need to knock off disparity_deletion_in[1] bases
-                                if len(hgvs_not_delins.posedit.edit.alt) == len(self.tx_hgvs_not_delins.posedit.edit.alt):
+                                if len(hgvs_not_delins.posedit.edit.alt) == len(
+                                        self.tx_hgvs_not_delins.posedit.edit.alt):
                                     if self.orientation == 1:
                                         self.tx_hgvs_not_delins.posedit.edit.ref = hgvs_not_delins.posedit.ref
                                     else:
@@ -2815,6 +2816,7 @@ def transcript_disparity(self, reverse_normalized_hgvs_genomic, stored_hgvs_not_
 
             if self.tx_hgvs_not_delins.posedit.pos.start.offset == 0 and \
                     self.tx_hgvs_not_delins.posedit.pos.end.offset == 0:
+
                 # In this instance, we have identified a transcript gap but the n. version of
                 # the transcript variant but do not have a position which actually hits the gap,
                 # so the variant likely spans the gap, and is not picked up by an offset.
@@ -2833,6 +2835,15 @@ def transcript_disparity(self, reverse_normalized_hgvs_genomic, stored_hgvs_not_
                     hgvs_refreshed_variant = self.tx_hgvs_not_delins
                     return hgvs_refreshed_variant
 
+                elif (((g3.posedit.pos.end.base - g3.posedit.pos.start.base) >
+                        (hgvs_genomic_norm.posedit.pos.end.base - hgvs_genomic_norm.posedit.pos.start.base)) and
+                        hgvs_genomic_norm.posedit.edit.type == 'del' and
+                        g3.posedit.pos.start.base < hgvs_genomic_norm.posedit.pos.start.base and
+                        (g3.posedit.pos.end.base + int(self.disparity_deletion_in[1])) <
+                      hgvs_genomic_norm.posedit.pos.end.base):
+                    hgvs_refreshed_variant = self.tx_hgvs_not_delins
+                    return hgvs_refreshed_variant
+
                 g3.posedit.pos.end.base = g3.posedit.pos.start.base + (len(g3.posedit.edit.ref) - 1)
                 try:
                     c2 = self.validator.vm.g_to_t(g3, c1.ac, alt_aln_method=self.validator.alt_aln_method)

VariantValidator/modules/gene2transcripts.py

@@ -6,6 +6,7 @@
 from . import utils as fn
 from . import seq_data
 
+
 # Pre compile variables
 vvhgvs.global_config.formatting.max_ref_length = 1000000
 
@@ -53,11 +54,15 @@ def gene2transcripts(g2t, query, validator=False, bypass_web_searches=False, sel
             query = g2t.db.get_stable_gene_id_from_hgnc_id(query)[1]
             if query == "No data":
                 try:
-                    query = validator.db.get_transcripts_from_annotations(store_query)
-                    for tx in query:
-                        if tx[5] != "unassigned":
-                            query = tx[5]
-                            break
+                    query = g2t.db.get_transcripts_from_annotations(store_query)
+                    if "none" not in query[0]:
+                        for tx in query:
+                            if tx[5] != "unassigned":
+                                query = tx[5]
+                                break
+                    else:
+                        return {'error': 'Unable to recognise HGNC ID. Please provide a gene symbol',
+                                "requested_symbol": store_query}
                 except TypeError:
                     pass
 
@@ -170,10 +175,21 @@ def gene2transcripts(g2t, query, validator=False, bypass_web_searches=False, sel
                     hgnc_id = vvta_record[0][0]
                     previous_sym = hgnc
                     symbol_identified = True
-                if len(vvta_record) > 1:
+                elif len(vvta_record) > 1:
                     return {'error': '%s is a previous symbol for %s genes. '
-                                     'Refer to https://www.genenames.org/' % (current_sym, str(len(vvta_record))),
+                                     'Refer to https://www.genenames.org/' % (hgnc, str(len(vvta_record))),
                             "requested_symbol": query}
+                else:
+                    # Is it an updated symbol?
+                    old_symbol = g2t.db.get_uta_symbol(hgnc)
+                    if old_symbol is not None:
+                        vvta_record = g2t.hdp.get_gene_info(old_symbol)
+                        if vvta_record is not None:
+                            current_sym = hgnc
+                            gene_name = vvta_record[3]
+                            hgnc_id = vvta_record[0]
+                            previous_sym = old_symbol
+                            symbol_identified = True
 
         if symbol_identified is False:
             return {'error': 'Unable to recognise gene symbol %s' % hgnc,

VariantValidator/modules/hgvs_utils.py

@@ -1058,7 +1058,7 @@ def hard_right_hgvs2vcf(hgvs_genomic, primary_assembly, hn, reverse_normalizer,
 
                 # Look for flank subs that may be missed when naieve mapping c > c made a delins from a sub
                 # This is a hgvs.py quirl for flanking subs in the antisense oriemntation and refers to
-                #
+                # https://github.com/openvar/variantValidator/issues/651
                 try:
                     normalized_end_seq_check_mapped = hn.normalize(end_seq_check_mapped)
                     normalized_end_seq_check_variant = hn.normalize(end_seq_check_variant)

tests/test_gene2transcript.py

@@ -185,6 +185,13 @@ def test_symbol_valid_hgnc_id(self):
         results = self.vv.gene2transcripts(symbol)
         assert results["current_symbol"] == "COL1A1"
 
+    def test_symbol_invalid_hgnc_id(self):
+        symbol = 'HGNC:12029'
+        results = self.vv.gene2transcripts(symbol)
+        print(results)
+        assert results["error"] == "Unable to recognise HGNC ID. Please provide a gene symbol"
+        assert results["requested_symbol"] == symbol
+
     def test_multiple_genes(self):
         symbol = '["HGNC:2197", "COL1A1", "2197"]'
         results = self.vv.gene2transcripts(symbol)