Alignment issue with NM_017680.6

[ifokkema]

Describe the bug
Originally reported by @mihailefter.
VariantValidator doesn't handle the mapping well from transcript to genome around a gap in NM_017680.6.

Bases NC_000009.12:g.92474743_92474745 are not present in NM_017680.6.
NC_000009.12:g.92474743_92474745del => NM_017680.6:c.151_153=.
Therefore, submitting NC_000009.12:g.92474742C>A returns NM_017680.6:c.153delinsTGAT on both the LOVD endpoint and the VV endpoint. This is as I expect.
However, when we submit NM_017680.6:c.153G>T (the "dumb" version of the genomic C>A) to the VV endpoint, the genomic mapping is NC_000009.12:g.92474742C>A, which is incorrect.

To Reproduce
See above.

Expected behavior
I expect to see NC_000009.12:g.92474742delinsATCA. However, when I submit that variant to the LOVD endpoint, I get NM_017680.6:c.153delinsTGAT back, the mapping earlier linked to the input NC_000009.12:g.92474742C>A. I had expected to see NM_017680.6:c.153G>T, instead.

Additional context
Notes: this gene is on the reverse strand, and the gap is within a repeat region. I suspect the gap isn't fully justified in both directions while applying the variant.

[Peter-J-Freeman]

I just want to break this down so we can write the relevant tests as well as tweak the code

import json
import VariantValidator
vval = VariantValidator.Validator()
variant = "NC_000009.12:g.92474743_92474745del"
genome_build = 'GRCh38'
select_transcripts = 'NM_017680.6'
transcript_set = 'refseq'
validate = vval.validate(variant, genome_build, select_transcripts, transcript_set)
validation = validate.format_as_dict(with_meta=True)
print(json.dumps(validation, sort_keys=True, indent=4, separators=(',', ': ')))

Gives

{
    "NM_017680.6:c.151_153=": {
        "alt_genomic_loci": [
            {
                "grch38": {
                    "hgvs_genomic_description": "NW_025791788.1:g.309472_309474=",
                    "vcf": {
                        "alt": "CTC",
                        "chr": "HG1012_PATCH",
                        "pos": "309472",
                        "ref": "CTC"
                    }
                }
            },
            {
                "hg38": {
                    "hgvs_genomic_description": "NW_025791788.1:g.309472_309474=",
                    "vcf": {
                        "alt": "CTC",
                        "chr": "NW_025791788.1",
                        "pos": "309472",
                        "ref": "CTC"
                    }
                }
            }
        ],
        "annotations": {
            "chromosome": "9",
            "db_xref": {
                "CCDS": null,
                "ensemblgene": null,
                "hgnc": "HGNC:14872",
                "ncbigene": "54829",
                "select": "MANE"
            },
            "ensembl_select": false,
            "mane_plus_clinical": false,
            "mane_select": true,
            "map": "9q22.31",
            "note": "asporin",
            "refseq_select": true,
            "variant": "1"
        },
        "gene_ids": {
            "ccds_ids": [],
            "ensembl_gene_id": "ENSG00000106819",
            "entrez_gene_id": "54829",
            "hgnc_id": "HGNC:14872",
            "omim_id": [
                "608135"
            ],
            "ucsc_id": "uc004ase.3"
        },
        "gene_symbol": "ASPN",
        "genome_context_intronic_sequence": "",
        "hgvs_lrg_transcript_variant": "",
        "hgvs_lrg_variant": "",
        "hgvs_predicted_protein_consequence": {
            "lrg_slr": "",
            "lrg_tlr": "",
            "slr": "NP_060150.4:p.(E51=)",
            "tlr": "NP_060150.4:p.(Glu51=)"
        },
        "hgvs_refseqgene_variant": "",
        "hgvs_transcript_variant": "NM_017680.6:c.151_153=",
        "primary_assembly_loci": {
            "grch37": {
                "hgvs_genomic_description": "NC_000009.11:g.95237066_95237068del",
                "vcf": {
                    "alt": "C",
                    "chr": "9",
                    "pos": "95237024",
                    "ref": "CTCA"
                }
            },
            "grch38": {
                "hgvs_genomic_description": "NC_000009.12:g.92474784_92474786del",
                "vcf": {
                    "alt": "C",
                    "chr": "9",
                    "pos": "92474742",
                    "ref": "CTCA"
                }
            },
            "hg19": {
                "hgvs_genomic_description": "NC_000009.11:g.95237066_95237068del",
                "vcf": {
                    "alt": "C",
                    "chr": "chr9",
                    "pos": "95237024",
                    "ref": "CTCA"
                }
            },
            "hg38": {
                "hgvs_genomic_description": "NC_000009.12:g.92474784_92474786del",
                "vcf": {
                    "alt": "C",
                    "chr": "chr9",
                    "pos": "92474742",
                    "ref": "CTCA"
                }
            }
        },
        "reference_sequence_records": {
            "protein": "https://www.ncbi.nlm.nih.gov/nuccore/NP_060150.4",
            "transcript": "https://www.ncbi.nlm.nih.gov/nuccore/NM_017680.6"
        },
        "refseqgene_context_intronic_sequence": "",
        "rna_variant_descriptions": null,
        "selected_assembly": "GRCh38",
        "submitted_variant": "NC_000009.12:g.92474743_92474745del",
        "transcript_description": "Homo sapiens asporin (ASPN), transcript variant 1, mRNA",
        "validation_warnings": [
            "Submitted description does not represent a true variant because it is an artefact of aligning NM_017680.6 with NC_000009.12 (genome build GRCh38)",
            "NM_017680.6 contains 3 fewer bases between c.152_153 than NC_000009.12"
        ],
        "variant_exonic_positions": {
            "NC_000009.12": {
                "end_exon": "2",
                "start_exon": "2"
            }
        }
    },
    "flag": "gene_variant",
    "metadata": {
        "variantvalidator_hgvs_version": "2.2.0",
        "variantvalidator_version": "2.2.1.dev729+g86e62d8.d20241105",
        "vvdb_version": "vvdb_2024_8",
        "vvseqrepo_db": "VV_SR_2024_09/master",
        "vvta_version": "vvta_2024_09"
    }
}

Which is correct

[Peter-J-Freeman]

import json
import VariantValidator
vval = VariantValidator.Validator()
variant = "NM_017680.6:c.153G>T"
genome_build = 'GRCh38'
select_transcripts = 'NM_017680.6'
transcript_set = 'refseq'
validate = vval.validate(variant, genome_build, select_transcripts, transcript_set)
validation = validate.format_as_dict(with_meta=True)
print(json.dumps(validation, sort_keys=True, indent=4, separators=(',', ': ')))

Is giving

{
    "NM_017680.6:c.153G>T": {
        "alt_genomic_loci": [
            {
                "grch38": {
                    "hgvs_genomic_description": "NW_025791788.1:g.309472C>A",
                    "vcf": {
                        "alt": "A",
                        "chr": "HG1012_PATCH",
                        "pos": "309472",
                        "ref": "C"
                    }
                }
            },
            {
                "hg38": {
                    "hgvs_genomic_description": "NW_025791788.1:g.309472C>A",
                    "vcf": {
                        "alt": "A",
                        "chr": "NW_025791788.1",
                        "pos": "309472",
                        "ref": "C"
                    }
                }
            }
        ],
        "annotations": {
            "chromosome": "9",
            "db_xref": {
                "CCDS": null,
                "ensemblgene": null,
                "hgnc": "HGNC:14872",
                "ncbigene": "54829",
                "select": "MANE"
            },
            "ensembl_select": false,
            "mane_plus_clinical": false,
            "mane_select": true,
            "map": "9q22.31",
            "note": "asporin",
            "refseq_select": true,
            "variant": "1"
        },
        "gene_ids": {
            "ccds_ids": [],
            "ensembl_gene_id": "ENSG00000106819",
            "entrez_gene_id": "54829",
            "hgnc_id": "HGNC:14872",
            "omim_id": [
                "608135"
            ],
            "ucsc_id": "uc004ase.3"
        },
        "gene_symbol": "ASPN",
        "genome_context_intronic_sequence": "",
        "hgvs_lrg_transcript_variant": "",
        "hgvs_lrg_variant": "",
        "hgvs_predicted_protein_consequence": {
            "lrg_slr": "",
            "lrg_tlr": "",
            "slr": "NP_060150.4:p.(E51D)",
            "tlr": "NP_060150.4:p.(Glu51Asp)"
        },
        "hgvs_refseqgene_variant": "",
        "hgvs_transcript_variant": "NM_017680.6:c.153G>T",
        "primary_assembly_loci": {
            "grch37": {
                "hgvs_genomic_description": "NC_000009.11:g.95237024C>A",
                "vcf": {
                    "alt": "A",
                    "chr": "9",
                    "pos": "95237024",
                    "ref": "C"
                }
            },
            "grch38": {
                "hgvs_genomic_description": "NC_000009.12:g.92474742C>A",
                "vcf": {
                    "alt": "A",
                    "chr": "9",
                    "pos": "92474742",
                    "ref": "C"
                }
            },
            "hg19": {
                "hgvs_genomic_description": "NC_000009.11:g.95237024C>A",
                "vcf": {
                    "alt": "A",
                    "chr": "chr9",
                    "pos": "95237024",
                    "ref": "C"
                }
            },
            "hg38": {
                "hgvs_genomic_description": "NC_000009.12:g.92474742C>A",
                "vcf": {
                    "alt": "A",
                    "chr": "chr9",
                    "pos": "92474742",
                    "ref": "C"
                }
            }
        },
        "reference_sequence_records": {
            "protein": "https://www.ncbi.nlm.nih.gov/nuccore/NP_060150.4",
            "transcript": "https://www.ncbi.nlm.nih.gov/nuccore/NM_017680.6"
        },
        "refseqgene_context_intronic_sequence": "",
        "rna_variant_descriptions": null,
        "selected_assembly": "GRCh38",
        "submitted_variant": "NM_017680.6:c.153G>T",
        "transcript_description": "Homo sapiens asporin (ASPN), transcript variant 1, mRNA",
        "validation_warnings": [],
        "variant_exonic_positions": {
            "NC_000009.12": {
                "end_exon": "2",
                "start_exon": "2"
            }
        }
    },
    "flag": "gene_variant",
    "metadata": {
        "variantvalidator_hgvs_version": "2.2.0",
        "variantvalidator_version": "2.2.1.dev729+g86e62d8.d20241105",
        "vvdb_version": "vvdb_2024_8",
        "vvseqrepo_db": "VV_SR_2024_09/master",
        "vvta_version": "vvta_2024_09"
    }
}

Which is not correct. Should be a genomic delins of NC_000009.12:g.92474742delinsATCA

@ifokkema please confirm this is the first issue.

I plan to fix this first, then keep working through the rest of the issue. Need VV to be correct before we work on VF/LOVD since they use the VV mappings (all be it somewhat oddly on occasion!)

[ifokkema]

(...)
variant = "NC_000009.12:g.92474743_92474745del"
(...)

Gives

{
    "NM_017680.6:c.151_153=": {
        "primary_assembly_loci": {
            "grch37": {
                "hgvs_genomic_description": "NC_000009.11:g.95237066_95237068del"
            },
            "grch38": {
                "hgvs_genomic_description": "NC_000009.12:g.92474784_92474786del"
            }
        }
    }
}

Which is correct

I suddenly realize that NM_017680.6:c.151_153= is kind of arbitrary to indicate that a three-nucleotide deletion on the genome has no effect on the transcript. I mean, why indicate on the transcript that three nucleotides haven't changed?

If repeat (...)GATGATGATGAG is unchanged on the transcript (the genome has one GAT extra), why would VV indicate that the GAG is unchanged? That is what c.151_153 is. Why not indicate that the last AG is unchanged (the flanking bases of where the insertion would have otherwise occurred) or indicate that the entire GAT-repeat is unchanged, possibly with flanking bases?

When testing, I submitted a slightly longer variant, which incorrectly also returned NM_017680.6:c.151_153=.
So, something is fishy there, too.

        "hgvs_transcript_variant": "NM_017680.6:c.153G>T",
        "primary_assembly_loci": {
            "grch37": {
                "hgvs_genomic_description": "NC_000009.11:g.95237024C>A"
            },
            "grch38": {
                "hgvs_genomic_description": "NC_000009.12:g.92474742C>A"
            }
        }

Which is not correct. Should be a genomic delins of NC_000009.12:g.92474742delinsATCA

@ifokkema please confirm this is the first issue.

Yes, I confirm, the result should not be NC_000009.12:g.92474742delinsATCA.

[Peter-J-Freeman]

I suddenly realize that NM_017680.6:c.151_153= is kind of arbitrary to indicate that a three-nucleotide deletion on the genome has no effect on the transcript. I mean, why indicate on the transcript that three nucleotides haven't changed?

This is internal and. preference. We state the "tested" bases. We try not to return c.= because this may falsely report that the entire c. is unaffected. We are only talking about specific bases. It is HGVS compliant as I understand it so will not change.

Yes, I confirm, the result should not be NC_000009.12:g.92474742delinsATCA.

fixed this so far. Workinng on getting NC_000009.12:g.92474742delinsATCA back to NM_017680.6:c.153G>T at the moment

Will then look at incorrectly also returned NM_017680.6:c.151_153=..

Will be fixed in stages and reported back ready for testing by you @ifokkema . Please add additiional comments as required here

[ifokkema]

I suddenly realize that NM_017680.6:c.151_153= is kind of arbitrary to indicate that a three-nucleotide deletion on the genome has no effect on the transcript. I mean, why indicate on the transcript that three nucleotides haven't changed?

This is internal and. preference. We state the "tested" bases. We try not to return c.= because this may falsely report that the entire c. is unaffected. We are only talking about specific bases. It is HGVS compliant as I understand it so will not change.

Oh, I agree that c.= is bad, but my point is that these are not the tested bases... These are three (why three?) bases after the repeat sequence. Shouldn't the repeat sequence be considered as tested, perhaps with two flanking nucleotides?

fixed this so far. Workinng on getting NC_000009.12:g.92474742delinsATCA back to NM_017680.6:c.153G>T at the moment

Will then look at incorrectly also returned NM_017680.6:c.151_153=..

Will be fixed in stages and reported back ready for testing by you @ifokkema . Please add additiional comments as required here

Cool, thanks a lot!

[Peter-J-Freeman]

Oh, I agree that c.= is bad, but my point is that these are not the tested bases... These are three (why three?) bases after the repeat sequence. Shouldn't the repeat sequence be considered as tested, perhaps with two flanking nucleotides?

It covers the span of the region tested. In this case, the variant is on the flank, so call it position 1. The inserted/deleted bases are between the next 2 bases. So overall these are the c. bases tested to make the variant, i.e. compared to the genomic. We use it internally for testing as well, so very reluctant to re-write ~1000 tests :)

[ifokkema]

Oh, I agree that c.= is bad, but my point is that these are not the tested bases... These are three (why three?) bases after the repeat sequence. Shouldn't the repeat sequence be considered as tested, perhaps with two flanking nucleotides?

It covers the span of the region tested. In this case, the variant is on the flank, so call it position 1. The inserted/deleted bases are between the next 2 bases. So overall these are the c. bases tested to make the variant, i.e. compared to the genomic. We use it internally for testing as well, so very reluctant to re-write ~1000 tests :)

I'm afraid not... Sorry for my terrible handwriting and bad drawing, but this (see figure) is the current situation. VV indicates the sequence after the variable region is unchanged. Note that this is the output for the genomic variant that changes the genomic repeat region to match the one in the transcript. There is no flanking variant.

[Peter-J-Freeman]

import json
import VariantValidator
vval = VariantValidator.Validator()
variant = "NC_000009.12:g.92474742delinsATCA"
genome_build = 'GRCh38'
select_transcripts = 'NM_017680.6'
transcript_set = 'refseq'
validate = vval.validate(variant, genome_build, select_transcripts, transcript_set)
validation = validate.format_as_dict(with_meta=True)
print(json.dumps(validation, sort_keys=True, indent=4, separators=(',', ': ')))

{
    "NM_017680.6:c.153G>T": {
        "alt_genomic_loci": [
            {
                "grch38": {
                    "hgvs_genomic_description": "NW_025791788.1:g.309472C>A",
                    "vcf": {
                        "alt": "A",
                        "chr": "HG1012_PATCH",
                        "pos": "309472",
                        "ref": "C"
                    }
                }
            },
            {
                "hg38": {
                    "hgvs_genomic_description": "NW_025791788.1:g.309472C>A",
                    "vcf": {
                        "alt": "A",
                        "chr": "NW_025791788.1",
                        "pos": "309472",
                        "ref": "C"
                    }
                }
            }
        ],
        "annotations": {
            "chromosome": "9",
            "db_xref": {
                "CCDS": null,
                "ensemblgene": null,
                "hgnc": "HGNC:14872",
                "ncbigene": "54829",
                "select": "MANE"
            },
            "ensembl_select": false,
            "mane_plus_clinical": false,
            "mane_select": true,
            "map": "9q22.31",
            "note": "asporin",
            "refseq_select": true,
            "variant": "1"
        },
        "gene_ids": {
            "ccds_ids": [],
            "ensembl_gene_id": "ENSG00000106819",
            "entrez_gene_id": "54829",
            "hgnc_id": "HGNC:14872",
            "omim_id": [
                "608135"
            ],
            "ucsc_id": "uc004ase.3"
        },
        "gene_symbol": "ASPN",
        "genome_context_intronic_sequence": "",
        "hgvs_lrg_transcript_variant": "",
        "hgvs_lrg_variant": "",
        "hgvs_predicted_protein_consequence": {
            "lrg_slr": "",
            "lrg_tlr": "",
            "slr": "NP_060150.4:p.(E51D)",
            "tlr": "NP_060150.4:p.(Glu51Asp)"
        },
        "hgvs_refseqgene_variant": "",
        "hgvs_transcript_variant": "NM_017680.6:c.153G>T",
        "primary_assembly_loci": {
            "grch37": {
                "hgvs_genomic_description": "NC_000009.11:g.95237024delinsATCA",
                "vcf": {
                    "alt": "ATCA",
                    "chr": "9",
                    "pos": "95237024",
                    "ref": "C"
                }
            },
            "grch38": {
                "hgvs_genomic_description": "NC_000009.12:g.92474742delinsATCA",
                "vcf": {
                    "alt": "ATCA",
                    "chr": "9",
                    "pos": "92474742",
                    "ref": "C"
                }
            },
            "hg19": {
                "hgvs_genomic_description": "NC_000009.11:g.95237024delinsATCA",
                "vcf": {
                    "alt": "ATCA",
                    "chr": "chr9",
                    "pos": "95237024",
                    "ref": "C"
                }
            },
            "hg38": {
                "hgvs_genomic_description": "NC_000009.12:g.92474742delinsATCA",
                "vcf": {
                    "alt": "ATCA",
                    "chr": "chr9",
                    "pos": "92474742",
                    "ref": "C"
                }
            }
        },
        "reference_sequence_records": {
            "protein": "https://www.ncbi.nlm.nih.gov/nuccore/NP_060150.4",
            "transcript": "https://www.ncbi.nlm.nih.gov/nuccore/NM_017680.6"
        },
        "refseqgene_context_intronic_sequence": "",
        "rna_variant_descriptions": null,
        "selected_assembly": "GRCh38",
        "submitted_variant": "NC_000009.12:g.92474742delinsATCA",
        "transcript_description": "Homo sapiens asporin (ASPN), transcript variant 1, mRNA",
        "validation_warnings": [
            "Submitted description does not represent a true variant because it is an artefact of aligning NM_017680.6 with NC_000009.12 (genome build GRCh38)",
            "NM_017680.6 contains 3 fewer bases between c.152_153 than NC_000009.12"
        ],
        "variant_exonic_positions": {
            "NC_000009.12": {
                "end_exon": "2",
                "start_exon": "2"
            }
        }
    },
    "flag": "gene_variant",
    "metadata": {
        "variantvalidator_hgvs_version": "2.2.0",
        "variantvalidator_version": "2.2.1.dev709+g6340024",
        "vvdb_version": "vvdb_2024_8",
        "vvseqrepo_db": "VV_SR_2024_09/master",
        "vvta_version": "vvta_2024_09"
    }
}

Will look at that c.= position. Not sure if it is fixable if wrong

[Peter-J-Freeman]

Quickly recording this before I go to lunch.

If we map the direct position of the c. sub we get
NC_000009.12:g.92474742C>A
NM_017680.6:c.153G>T

Will help to check the c.= positioning

[Peter-J-Freeman]

I'm afraid not... Sorry for my terrible handwriting and bad drawing, but this (see figure) is the current situation. VV indicates the sequence after the variable region is unchanged. Note that this is the output for the genomic variant that changes the genomic repeat region to match the one in the transcript. There is no flanking variant.

Ah, I see why now. It is because the additionl bases are absent from the transcript, so the c. = position is the bases taken into consideration when making the mappings I believe. It is not particularly easy to generate to an exact position. Perhaps we can gather a few ecamples and tweak the code, but I'm gonna be upset at fixing the tests :)

[ifokkema]

Ah, I see why now. It is because the additionl bases are absent from the transcript, so the c. = position is the bases taken into consideration when making the mappings I believe. It is not particularly easy to generate to an exact position. Perhaps we can gather a few ecamples and tweak the code, but I'm gonna be upset at fixing the tests :)

I can imagine 😟
I don't think I can compensate for it from my side, as I can't tell what the full range is for the repeat... Well, if, somehow, it's easier to provide more information in VV's output about the mismatch between the genome and the transcript, in theory, LOVD could recalculate the "correct" c.?= notation from that. But, obviously, it would be better if VV could handle that natively.

[Peter-J-Freeman]

This can be worked on. Want to get the variants workig first, then look at some of the tests we have and see if I can tweak.

[Peter-J-Freeman]

@ifokkema

Does this look right?

import json
import VariantValidator
vval = VariantValidator.Validator()
variant = "NC_000009.12:g.92474743_92474748del"
genome_build = 'GRCh38'
select_transcripts = 'NM_017680.6'
transcript_set = 'refseq'
validate = vval.validate(variant, genome_build, select_transcripts, transcript_set)
validation = validate.format_as_dict(with_meta=True)
print(json.dumps(validation, sort_keys=True, indent=4, separators=(',', ': ')))

{
    "NM_017680.6:c.150_152del": {
        "alt_genomic_loci": [
            {
                "grch38": {
                    "hgvs_genomic_description": "NW_025791788.1:g.309511_309513del",
                    "vcf": {
                        "alt": "C",
                        "chr": "HG1012_PATCH",
                        "pos": "309472",
                        "ref": "CTCA"
                    }
                }
            },
            {
                "hg38": {
                    "hgvs_genomic_description": "NW_025791788.1:g.309511_309513del",
                    "vcf": {
                        "alt": "C",
                        "chr": "NW_025791788.1",
                        "pos": "309472",
                        "ref": "CTCA"
                    }
                }
            }
        ],
        "annotations": {
            "chromosome": "9",
            "db_xref": {
                "CCDS": null,
                "ensemblgene": null,
                "hgnc": "HGNC:14872",
                "ncbigene": "54829",
                "select": "MANE"
            },
            "ensembl_select": false,
            "mane_plus_clinical": false,
            "mane_select": true,
            "map": "9q22.31",
            "note": "asporin",
            "refseq_select": true,
            "variant": "1"
        },
        "gene_ids": {
            "ccds_ids": [],
            "ensembl_gene_id": "ENSG00000106819",
            "entrez_gene_id": "54829",
            "hgnc_id": "HGNC:14872",
            "omim_id": [
                "608135"
            ],
            "ucsc_id": "uc004ase.3"
        },
        "gene_symbol": "ASPN",
        "genome_context_intronic_sequence": "",
        "hgvs_lrg_transcript_variant": "",
        "hgvs_lrg_variant": "",
        "hgvs_predicted_protein_consequence": {
            "lrg_slr": "",
            "lrg_tlr": "",
            "slr": "NP_060150.4:p.(D50del)",
            "tlr": "NP_060150.4:p.(Asp50del)"
        },
        "hgvs_refseqgene_variant": "",
        "hgvs_transcript_variant": "NM_017680.6:c.150_152del",
        "primary_assembly_loci": {
            "grch37": {
                "hgvs_genomic_description": "NC_000009.11:g.95237063_95237068del",
                "vcf": {
                    "alt": "C",
                    "chr": "9",
                    "pos": "95237024",
                    "ref": "CTCATCA"
                }
            },
            "grch38": {
                "hgvs_genomic_description": "NC_000009.12:g.92474781_92474786del",
                "vcf": {
                    "alt": "C",
                    "chr": "9",
                    "pos": "92474742",
                    "ref": "CTCATCA"
                }
            },
            "hg19": {
                "hgvs_genomic_description": "NC_000009.11:g.95237063_95237068del",
                "vcf": {
                    "alt": "C",
                    "chr": "chr9",
                    "pos": "95237024",
                    "ref": "CTCATCA"
                }
            },
            "hg38": {
                "hgvs_genomic_description": "NC_000009.12:g.92474781_92474786del",
                "vcf": {
                    "alt": "C",
                    "chr": "chr9",
                    "pos": "92474742",
                    "ref": "CTCATCA"
                }
            }
        },
        "reference_sequence_records": {
            "protein": "https://www.ncbi.nlm.nih.gov/nuccore/NP_060150.4",
            "transcript": "https://www.ncbi.nlm.nih.gov/nuccore/NM_017680.6"
        },
        "refseqgene_context_intronic_sequence": "",
        "rna_variant_descriptions": null,
        "selected_assembly": "GRCh38",
        "submitted_variant": "NC_000009.12:g.92474743_92474748del",
        "transcript_description": "Homo sapiens asporin (ASPN), transcript variant 1, mRNA",
        "validation_warnings": [
            "Submitted description does not represent a true variant because it is an artefact of aligning NM_017680.6 with NC_000009.12 (genome build GRCh38)",
            "NM_017680.6 contains 3 fewer bases between c.152_153 than NC_000009.12"
        ],
        "variant_exonic_positions": {
            "NC_000009.12": {
                "end_exon": "2",
                "start_exon": "2"
            }
        }
    },
    "flag": "gene_variant",
    "metadata": {
        "variantvalidator_hgvs_version": "2.2.0",
        "variantvalidator_version": "2.2.1.dev709+g6340024",
        "vvdb_version": "vvdb_2024_8",
        "vvseqrepo_db": "VV_SR_2024_09/master",
        "vvta_version": "vvta_2024_09"
    }
}

So, the deletion g. normalises 3 prime. Spans the gap. The deletion in the g. is 6 bases. There are 3 bases less in the transcript. we seem to get from the g. to the c. and back to the g.