No variant summary was recorded for this case.
Five-step scale from benign to pathogenic. This variant is uncertain. Select a step to read what it means.
Uncertain.The recorded classification is uncertain significance.
Gene-level background from the recorded interpretation.
Words with a dotted underline have a short definition. Hover or tap them.GALNT12 encodes a member of the UDP-GalNAc:polypeptide N-acetylgalactosaminyltransferase family, enzymes that initiate O-linked glycosylation by transferring N-acetylgalactosamine (GalNAc) to serine or threonine residues on proteins. These glycosylation modifications are important for the proper function of many cellular proteins. Mutations in this gene are associated with increased susceptibility to colorectal cancer, indicating a role in colorectal cancer predisposition.
The recorded framework is Not recorded. The criteria and strengths shown below come from the saved assessment; this page does not recalculate the classification.
The applied criteria and their strengths are shown as assessed under the stated framework.
No pathogenic criterion marked as applied in the recorded assessment.
Met at supporting strength: SpliceAI max delta 0.017 is at or below the <=0.1 no-splice-impact BP4 cutoff, counted once and not reused for BP7.
VUS: BP4 (supporting, SpliceAI max delta 0.017 indicating no splice impact) is the sole met criterion, below the two-supporting-benign threshold for Likely Benign.
Evidence summaries and captured source records, where available. Missing entries should not be interpreted as negative findings.
gnomAD v4.1: 733 / 1,611,402 alleles. Recorded AF 0.000454883. Recorded homozygote count 1.
This variant is present in gnomAD v2.1 (AF= 0.00073676; MAF= 0.07368%, 207/280960 alleles, homozygotes = 0) and has highest observed frequency in the European (non-Finnish) population (AF= 0.00134836; MAF= 0.13484%, 173/128304 alleles, homozygotes = 0); grpmax FAF= 0.00117196.
This variant is present in gnomAD-Canada v1.0 (AF= 0.0005992590978426672, 11/18356 alleles, homozygotes = 0).
Recorded SpliceAI maximum delta score: 0.017. Refer to the recorded PP3/BP4 rationale for interpretation.
4 Likely benign · 2 Benign. ClinVar API review status: criteria provided, multiple submitters, no conflicts.
No standalone functional evidence summary is recorded in evidence_summary.functional.
This variant has previously been reported in somatic cancers (COSMIC; COSV100899069, n = 1 times).
This variant does not lie in a statistically significant hotspot.
ACMG/AMP criteria under Not recorded, with the data and sources behind each one.
28 criteria with recorded statuses: 1 met · 8 not met · 11 not assessed · 8 not applicable. Open any row for the full reasoning, the rule and the evidence behind it.
All four segregation/de novo criteria (PS2, PM6, PP1, BS4) are not assessed because the case contains no proband, no parental genotypes, no pedigree, and no meioses for NM_024642.5:c.567T>C; the available evidence is variant-level database annotation only.
No GALNT12 VCEP or ClinGen CSPEC specification was retrieved (framework_complete=false), so the generic ACMG/AMP 2015 criteria of Richards et al. (PMID:25741868) were used for every criterion in this group.
This group therefore contributes no pathogenic or benign evidence, and the absence of segregation/de novo data was not converted into a benign inference.
Framework selection: no ClinGen CSPEC or gene-specific VCEP specification exists for GALNT12 (vcep_materials.json: no cspec url available; case_summary.json cspec found=false; final_classification_framework source='generic_acmg_fallback'), so the generic ACMG/AMP 2015 criteria govern this group.
No criterion in this group is met. PP5 and BP6 are definitively not met because ClinVar VCV000416212 (exact transcript-HGVS match for c.567T>C) has expert_panel_submissions=0 and is_expert_panel=false for all six submissions, with an aggregate 2-star multi-submitter status that cannot substitute for an expert-panel assertion.
PS4, PP4 and BP5 are not_assessed for lack of the case-level data they require: no case-control or affected-carrier data (PS4), no proband phenotype or HPO terms (PP4), and no proband with an alternate molecular basis for disease (BP5).
Numeric-threshold direction check: none of the five criteria in this group has an evaluable numeric threshold in this case. No VCEP specification governs GALNT12, and the generic ACMG/AMP framework and the supplied published-calibration block define no numeric cutoff for PS4, PP4, PP5, BP5 or BP6, so no comparison operator exists to be read or misapplied.
This group covers the allele-level criteria PM3 and BP2, both of which require knowledge of how the variant is phased relative to another pathogenic allele; neither can be evaluated from database-level evidence alone.
No allele-level data exist for this case: there is no proband genotype, pedigree, segregation, haplotype, or parental-origin information anywhere in the case bundle or in the remaining case output files, and none of the six ClinVar submissions for variation 416212 records a co-occurring variant or a cis/trans relationship.
The variant is NM_024642.5(GALNT12):c.567T>C (NP_078918.3:p.(Asn189=)), a synonymous change; GALNT12 has an empty documented inheritance list and is described as a dominant colorectal cancer susceptibility gene, so even the recessive-disorder prerequisite for PM3 cannot be established.
GALNT12 has no ClinGen CSPEC and no VCEP specification (framework_complete=false; generic ACMG/AMP 2015 fallback), so the generic definitions of PM3 and BP2 from PMID:25741868 were applied.
Both criteria are returned as not_assessed: PM3 contributes no pathogenic-direction weight and BP2 contributes no benign-direction weight to the final classification. Any downstream classification must be driven by the criteria in the other groups.
This group owns the calibrated computational-evidence criteria PP3, BP4 and BP7. All three were scoped against the variant type before any predictor score was read: NM_024642.5:c.567T>C is a synonymous substitution (NP_078918.3:p.(Asn189=)), so PP3/BP4 are evaluated on the SpliceAI splice path only and BP7 is the synonymous-variant criterion; the REVEL missense path and BayesDel are not applicable to this variant.
The group's only positive contribution to classification is BP4 at supporting strength, from the SpliceAI maximum delta score of 0.017 for this variant (at or below the <= 0.1 no-splice-impact cutoff; Jaganathan et al. 2019, Cell, PMID:30661751). PP3 is not met (0.017 is far below the >= 0.2 supporting cutoff).
BP7 is not counted: the single SpliceAI/Pangolin no-impact prediction that would support it is already credited under BP4, and one prediction may not be double-counted across PP3/BP4/BP7; the conservation element of BP7 is also unverified (no phyloP/GERP/PhastCons data in the case).
Net contribution: one benign supporting criterion (BP4, supporting) and no pathogenic-direction computational evidence. No VCEP/CSPEC specification for GALNT12 was available, so generic ACMG/AMP 2015 in-silico calibration was applied throughout.
No ClinGen CSPEC/VCEP specification was retrieved for GALNT12 (gene_vcep_dir null; VCEP materials record 'no cspec url available'), so all four population criteria were adjudicated under the generic ACMG/AMP 2015 framework with generic published thresholds and the all-comers gnomAD totals as the default source; the non-cancer subsets (gnomAD v2.1 non-cancer exomes, gnomAD v3.1 non-cancer genomes) and gnomAD-Canada were reviewed as corroboration and neither changes any verdict nor is mandated by any specification for this gene.
NM_024642.5:c.567T>C (GRCh37 9-101589059-T-C / GRCh38 9-98826777-T-C, p.(Asn189=)) is a low-frequency but clearly present variant: 0.074% in gnomAD v2.1 (207/280,960), 0.045% in gnomAD v4.1 (733/1,611,402), 0.069% and 0.053% in the v2.1 and v3.1 non-cancer subsets, and 0.060% in gnomAD-Canada, with the highest ancestry-specific frequency 0.151% in the v3.1 non-cancer Ashkenazi Jewish group.
Against generic thresholds (BA1 >= 0.05, BS1 >= 0.01, PM2 <= 0.0001), the variant fails both benign stand-alone/strong frequency criteria by 1-2 orders of magnitude and equally fails PM2: it is far too common to be treated as rare (4.5-15 fold above 0.0001 depending on dataset and ancestry) yet far too rare for a frequency-based benign call (BS1 missed by ~6.6-7.4 fold at the highest ancestry-specific frequencies). The net population picture is 'not rare, but not common enough for BS1/BA1'.
Homozygote review: no homozygote exists in gnomAD v2.1, either non-cancer subset, or gnomAD-Canada; gnomAD v4.1 reports exactly one homozygote in a European (non-Finnish) XX individual. Because GALNT12-associated disease is adult-onset, incompletely penetrant colorectal cancer susceptibility, this observation does not support BS2 (which requires full penetrance expected at an early age) and is retained for review.
No criteria match.
Scope first, identical determination as for PP3: NM_024642.5:c.567T>C is a synonymous substitution (NP_078918.3:p.(Asn189=)) in exon 3 of GALNT12, so the applicable BP4 path is the splice-prediction path (SpliceAI); there is no REVEL protein-impact path to use because a synonymous change has no amino-acid substitution to score, and the bundle's REVEL result is not found. Because BP4 is being taken from one path only, the missense-path score is neither used nor needed. SpliceAI Lookup for this exact variant gives a maximum delta score of 0.017 (DS_DG 0.017 is the largest of DS_AG 0.000 / DS_AL 0.007 / DS_DG 0.017 / DS_DL 0.007), with Pangolin splice-gain 0.01 and splice-loss -0.005. That maximum delta score is at or below the recommended SpliceAI no-impact cutoff of <= 0.1, which is the published BP4 (supporting) calibration for splice predictors (Jaganathan et al. 2019, Cell; PMID:30661751). Two independent directions agree that no splice consequence is predicted: no donor or acceptor is gained or strengthened (largest gain 0.017, at 21 bp downstream), and the acceptor-loss / donor-loss components are 0.000 and 0.007. The result is also internally consistent with the variant's position - c.567 sits 25 bp inside exon 3 (exon 3 spans c.542-731), i.e. away from the exon 3 acceptor (DP_AG -25) - and no cryptic splice site usage is predicted. BP4 is therefore met at supporting strength: multiple in-silico lines of splice evidence indicate no impact on splicing, which for a synonymous change is the relevant gene-product-level claim. This single splice prediction is credited here only; it is not also counted under BP7 (see BP7), and the framework's instruction that PP3/BP4 evidence come from one path only is respected. No VCEP/CSPEC specification for GALNT12 was available (cspec found=false; generic ACMG/AMP fallback, framework_complete=false), so the generic published threshold was applied verbatim.
Generic ACMG/AMP in-silico splice path: SpliceAI max delta <= 0.1 -> BP4 (supporting) (Jaganathan et al. 2019, Cell; PMID:30661751). Met: max delta 0.017.
PM2 (supporting) requires the variant to be absent from, or extremely rare in, population databases, operationalized here at <= 0.0001 (ClinGen SVI recommendation to downgrade PM2 to supporting; PMID:25741868). No GALNT12 VCEP/CSPEC specification was retrieved, so generic thresholds and the all-comers gnomAD totals, with the non-cancer subsets as corroboration, were used. The variant is not rare at that level in any dataset: gnomAD v2.1 all-comers AF 0.00073676 (207/280,960 alleles) and gnomAD v4.1 all-comers AF 0.000454883 (733/1,611,402 alleles) are both above 0.0001; the non-cancer subsets are likewise above it (v2.1 non-cancer exomes 0.000688635, 162/235,248; v3.1 non-cancer genomes 0.000534347, 79/147,844); gnomAD-Canada is 0.000599259 (11/18,356). Using the highest ancestry-specific value, grpmax FAF is 0.00117196 in v2.1 (European non-Finnish AF 0.00134836) and grpmax FAF 0.00048901 in v4.1, while the v3.1 non-cancer Ashkenazi Jewish subset reaches 0.00151423 - roughly 4.5- to 15-fold above the 0.0001 threshold. Because PM2 is met only when the variant is absent or at most 0.0001 in frequency, and it exceeds that in the combined cohorts, in every non-cancer subset, and in every major ancestry examined, PM2 is not met. This is a genuine frequency excess rather than a threshold artifact: the counts are allele-specific for the C allele at a split multi-allelic site, and the variant is observed in multiple independent cohorts (gnomAD v2.1, v4.1, v3.1 non-cancer and gnomAD-Canada), so it cannot be attributed to one population isolate or one dataset.
generic SVI population threshold: PM2 supporting <= 0.0001 (PMID:25741868)
Scope first. The variant is NM_024642.5:c.567T>C, which VariantValidator/Mutalyzer normalise to NP_078918.3:p.(Asn189=) (AAT>AAC, codon 189), i.e. a synonymous substitution in exon 3 of GALNT12 (chromosome 9q22.33; NC_000009.11:g.101589059T>C / NC_000009.12:g.98826777T>C) - not a missense change, and not a frameshift, nonsense or other protein-length-changing variant. PP3 for this variant therefore comes from the splice-prediction path (SpliceAI) only: there is no missense protein-impact claim to make for a synonymous change, and the REVEL protein-impact path is not applicable (the case bundle records REVEL as not found, and a missing missense score is not evidence and is not a reason to reach for a splice score in reverse). BayesDel is likewise not usable: no generic ACMG-strength calibration exists for it and no VCEP specification supplies one for GALNT12. No ClinGen CSPEC or gene-specific VCEP specification exists for GALNT12 (cspec found=false; final classification framework resolved to the generic ACMG/AMP fallback, framework_complete=false), so the generic ACMG/ClinGen-SVI calibration governs. SpliceAI Lookup for this exact variant reports DS_AG 0.000, DS_AL 0.007, DS_DG 0.017, DS_DL 0.007, giving a maximum delta score of 0.017 (the source's own summary sentence rounds this to 0.02); the nearest predicted cryptic acceptor is 25 bp upstream (DP_AG -25) and the nearest cryptic donor 21 bp downstream (DP_DG +21), i.e. neither is created or strengthened at the variant. Pangolin, reported alongside, gives splice-gain 0.01 and splice-loss -0.005. A maximum delta score of 0.017 is far below the recommended SpliceAI supporting cutoff for PP3 (max delta >= 0.2) and below the high-precision moderate cutoff (>= 0.5) (Jaganathan et al. 2019, Cell; PMID:30661751). No in-silico tool predicts a splice-altering effect, and no other computational evidence exists in this case that would support a damaging impact. PP3 is therefore not met. The variant's exonic position 25 bp into exon 3 (exon 3 spans c.542-731) is consistent with the absent splice signal, but PP3 rests on the predictor result, not on the location alone.
Generic ACMG/AMP in-silico splice path: SpliceAI max delta >= 0.2 -> PP3 (supporting); >= 0.5 -> PP3 (moderate) (Jaganathan et al. 2019, Cell; PMID:30661751). Not met: max delta 0.017.
PP5 is restricted to an exact-variant classification issued by a recognised expert panel: only an expert-panel (ClinGen VCEP-type) assertion of Pathogenic or Likely pathogenic for this specific variant may trigger it, while ordinary laboratory submissions and aggregate ClinVar labels must never do so. Direct check of the exact variant's ClinVar record VCV000416212 (NM_024642.5(GALNT12):c.567T>C, p.Asn189=; matched by transcript HGVS with exact_match=true) shows six germline submissions, all from clinical testing laboratories: GeneDx (SCV002028417, Likely benign), Ambry Genetics (SCV000673517, Likely benign), PreventionGenetics/Exact Sciences (SCV004728987, Likely benign, no assertion criteria), Sinai Health System (SCV006056651, Likely benign), Labcorp Genetics/formerly Invitae (SCV000561582, Benign) and Myriad Genetics (SCV007646811, Benign). Every one of these submissions carries is_expert_panel=false, and the submission audit counts expert_panel_submissions=0; the source_data expert_panels object is empty. The aggregate record status is 'criteria provided, multiple submitters, no conflicts' (2 stars), which is an ordinary multi-laboratory consensus, not a 3-star expert-panel assertion. In addition, every submission is benign-leaning, so even setting the expert-panel question aside there is no Pathogenic/Likely pathogenic assertion for this variant to serve as the substrate for PP5. Result of the check: expert-panel Pathogenic/Likely pathogenic classification of c.567T>C present? No. PP5 is therefore not met. No numeric threshold applies to this criterion (it is assertion-based, and no GALNT12 VCEP specification exists), so no comparison operator or inequality is required.
PP5 (supporting): the exact variant has been classified as Pathogenic or Likely pathogenic by a recognised expert panel (e.g. ClinGen VCEP) in a reputable source. Ordinary laboratory submissions, aggregate ClinVar labels, and non-expert assertions must not trigger PP5.
BA1 (stand-alone benign) requires an allele frequency >= 0.05 (Richards et al. 2015, PMID:25741868). No ClinGen CSPEC/VCEP specification was retrieved for GALNT12 (gene_vcep_dir is null; the VCEP materials record states 'no cspec url available'), so the generic ACMG/AMP 2015 thresholds apply and no non-cancer or exome-only population source is mandated for this gene; the all-comers gnomAD totals are therefore the default, with the non-cancer subsets reviewed as corroboration. Every frequency observed for NM_024642.5:c.567T>C (GRCh37 9-101589059-T-C / GRCh38 9-98826777-T-C) sits more than an order of magnitude below 0.05: gnomAD v2.1 all-comers total AF 0.00073676 (207/280,960 alleles, 0 homozygotes) with grpmax FAF 0.00117196; gnomAD v4.1 all-comers total AF 0.000454883 (733/1,611,402 alleles, 1 homozygote) with grpmax FAF 0.00048901; gnomAD-Canada v1.0 AF 0.000599259 (11/18,356 alleles, 0 homozygotes). The highest ancestry-specific frequency seen in any dataset, which is the value that would drive a stand-alone benign call, is 0.00151423 in the gnomAD v3.1 non-cancer Ashkenazi Jewish subset (5/3,302) and 0.00134836 in gnomAD v2.1 European (non-Finnish) (173/128,304) - approximately 33- to 37-fold below the 5% threshold. Because the criterion cannot be met even using the most enriched ancestry rather than the combined cohort, BA1 is not met. Threshold artifacts were explicitly checked: this position is a split multi-allelic site in the source records (was_split true, n_alt_alleles 2), so all figures used are allele-specific counts for the C allele rather than site-level frequencies, and the presence of a homozygote in v4.1 does not alter the allele-frequency comparison.
generic ACMG/AMP 2015 population threshold: BA1 stand-alone benign >= 0.05 (PMID:25741868)
BS1 (strong benign) requires an allele frequency greater than the disease prevalence, defaulted here to >= 0.01 (generic SVI default, per-gene derivation preferred; Whiffin et al. 2017, PMID:28518168). No GALNT12 VCEP/CSPEC specification was retrieved, and no prevalence-derived GALNT12-specific BS1 threshold has been published, so the generic 0.01 value governs and the all-comers gnomAD totals are the default source. All observed frequencies fall below 0.01: gnomAD v2.1 all-comers AF 0.00073676 (207/280,960), gnomAD v4.1 all-comers AF 0.000454883 (733/1,611,402), gnomAD v2.1 non-cancer exomes 0.000688635 (162/235,248), gnomAD v3.1 non-cancer genomes 0.000534347 (79/147,844), gnomAD-Canada 0.000599259 (11/18,356). The largest ancestry-specific value in any dataset is 0.00151423 (gnomAD v3.1 non-cancer Ashkenazi Jewish, 5/3,302), about 6.6-fold below 0.01; the largest all-comers subpopulation value is 0.00134836 (v2.1 European non-Finnish, 173/128,304), about 7.4-fold below. Because the criterion is evaluated against the total cohort and, for ancestry, against the highest-frequency subpopulation, and both remain below the threshold, BS1 is not met. The result is not marginal to the extent of changing the verdict, but it is close enough in the European and Ashkenazi Jewish ancestries (0.13-0.15%) that a formally derived, prevalence-based gene-specific threshold for a reduced-penetrance colorectal cancer predisposition gene could in principle be lower than 1%; no such published threshold exists, so the generic default is retained and the point is recorded as a residual gap rather than used to meet the criterion.
generic ACMG/AMP population threshold: BS1 strong benign >= 0.01 (PMID:28518168)
BS2 requires observation of the variant in a healthy adult for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder with full penetrance expected at an early age (Richards et al. 2015, PMID:25741868). The gene context for GALNT12 is increased susceptibility to colorectal cancer (OMIM 610290; NCBI Gene 79695), an adult-onset, incompletely penetrant, dominantly acting predisposition, so neither heterozygote nor homozygote observations in a population database can satisfy BS2's early-onset/full-penetrance precondition, which is the criterion's structural prerequisite. Factually, the variant is carried by 207/280,960 alleles in gnomAD v2.1 (0 homozygotes) and 733/1,611,402 alleles in gnomAD v4.1, where exactly one homozygote is reported (exome homozygote count 1, in a European non-Finnish XX individual); both non-cancer subsets show zero homozygotes (v2.1 non-cancer exomes 162/235,248; v3.1 non-cancer genomes 79/147,844), as does gnomAD-Canada (11/18,356). The single v4.1 homozygote therefore is the only homozygous observation anywhere, and it does not establish a healthy adult homozygous carrier for an early-onset fully penetrant condition; if anything, surviving homozygosity at 0.05% carrier frequency argues against a fully penetrant early-onset recessive model. gnomAD also aggregates cohorts that are not health-screened, so even the heterozygous carriers cannot be asserted to be unaffected healthy adults. BS2 is therefore not met, with the homozygote finding recorded for review.
generic ACMG/AMP 2015 BS2: observed in a healthy adult for a disorder with full penetrance expected at an early age (PMID:25741868)
BP6 is the benign counterpart of PP5 and is likewise restricted to an exact-variant classification issued by a recognised expert panel: only an expert-panel (ClinGen VCEP-type) assertion of Benign or Likely benign for this specific variant may trigger it. Ordinary laboratory submissions and aggregate ClinVar labels must never do so, even when they agree with one another. Direct check of the exact variant's ClinVar record VCV000416212 (NM_024642.5(GALNT12):c.567T>C, p.Asn189=, transcript-HGVS exact match) shows six germline submissions, all from clinical testing laboratories: GeneDx (Likely benign), Ambry Genetics (Likely benign), PreventionGenetics/Exact Sciences (Likely benign), Sinai Health System (Likely benign), Labcorp Genetics/formerly Invitae (Benign) and Myriad Genetics (Benign). Every submission has is_expert_panel=false and the submission audit records expert_panel_submissions=0; the ClinVar source_data expert_panels object is empty. The aggregate record status is 'criteria provided, multiple submitters, no conflicts' (2 stars) with an aggregate label of Benign/Likely benign - a consensus of independent clinical laboratories, not a single expert-panel classification, and therefore explicitly ineligible to trigger BP6. Result of the check: expert-panel Benign/Likely benign classification of c.567T>C present? No. BP6 is therefore not met. The concordance of six benign-leaning laboratory submissions is noted for context only and was not scored here; the population-frequency and in-silico groups remain the correct home for the substantive benign evidence in this case.
BP6 (supporting): the exact variant has been classified as Benign or Likely benign by a recognised expert panel (e.g. ClinGen VCEP) in a reputable source. Ordinary laboratory submissions, aggregate ClinVar labels, and non-expert assertions must not trigger BP6.
Scope: BP7 applies only to a synonymous variant, and NM_024642.5:c.567T>C is synonymous (NP_078918.3:p.(Asn189=), AAT>AAC at codon 189), so BP7 is in scope and was evaluated rather than dismissed. BP7 as defined in the ACMG/AMP 2015 framework (Richards et al., PMID:25741868) requires a synonymous (silent) variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site, and for which the nucleotide is not highly conserved. The case contains exactly one item of in-silico evidence that speaks to the first requirement: the SpliceAI result for this variant - maximum delta score 0.017, with Pangolin splice-gain 0.01 and splice-loss -0.005 - which does predict no splice impact. That same single prediction, however, is the evidence already credited to BP4 at supporting strength for this variant (SpliceAI max delta <= 0.1). The ACMG/AMP framework states that predictions from different in-silico tools are treated as a single piece of evidence rather than independent pieces of evidence, and this framework additionally forbids double-counting one prediction across PP3/BP4/BP7; crediting the identical SpliceAI/Pangolin result under both BP4 and BP7 would count one prediction twice and inflate the benign evidence for this variant. The second BP7 element cannot be established either: no nucleotide-conservation metric for c.567 (phyloP, GERP, PhastCons or equivalent) is present anywhere in the case evidence, so the 'not highly conserved' clause of BP7 is unverified. BP7 is therefore not counted as an independent benign criterion for this variant; the prediction that would support it has already been applied once, under BP4.
BP7 (Richards et al. 2015, PMID:25741868): synonymous variant with no predicted impact on the splice consensus sequence or creation of a new splice site, and the nucleotide not highly conserved. Not credited here because the only supporting prediction is already counted under BP4 and the conservation element is unverified.
PS2 requires a variant observed to have arisen de novo, with both parental samples shown through identity testing to be the biological parents and with a phenotype matching the gene's disease association with reasonable specificity (Richards et al. 2015, PMID:25741868). For NM_024642.5:c.567T>C (GALNT12, NP_078918.3:p.(Asn189=)) the case evidence is variant-level only: no proband, no parental genotypes, no trio results, no identity/kinship confirmation, and no phenotype information were retrieved. De novo origin therefore cannot be established and none of PS2's required conditions can be evaluated. No GALNT12 VCEP or ClinGen CSPEC specification was available (framework_complete=false), so the generic ACMG/AMP 2015 framework applies. PS2 is not met on the available evidence, and is recorded as not assessed because the necessary proband and parental data are simply absent rather than negative.
ACMG/AMP 2015 PS2 (Richards et al. 2015, PMID:25741868): requires confirmed de novo origin with both parental samples tested and identity-confirmed, plus a specific phenotype match
Not assessed. PS3 (Richards et al. 2015, PMID:25741868) requires a well-established in vitro or in vivo functional assay demonstrating a damaging effect on gene or protein function. The variant is NM_024642.5:c.567T>C in GALNT12, a synonymous substitution encoding NP_078918.3:p.(Asn189=); no CSPEC or gene-specific VCEP specification exists for GALNT12, and the case framework resolved to the generic ACMG/AMP 2015 fallback (framework_complete=false), so the generic PS3 definition applies unmodified. A sweep of every retrieved source found no functional assay of any kind for this variant: there is no in vitro GALNT12 enzyme-activity, minigene, RT-PCR/splicing, protein-stability, or reporter result in the source evidence, in the prefetch data, or in the literature pass. The only paper triaged to the 'functional' theme is PMID:25741868, which is the ACMG/AMP 2015 methodology paper itself and does not mention GALNT12, c.567T>C, or p.(N189=) (variant_mentioned=false); the medium-priority paper PMID:28492532 (Sherloc) is likewise a framework-refinement paper with no variant-level evidence. The only variant-level functional-adjacent datum is a computational splice prediction (SpliceAI max delta 0.017), which is in silico evidence and cannot substitute for a PS3 assay. The ClinVar record (ClinVarID 416212, Benign/Likely benign, five submitters) contains no variant-level assay, and its criterion-lead submissions cite only framework references (PMID:25741868, PMID:28492532). With no qualifying functional evidence available, PS3 cannot be assessed from this dataset.
Generic ACMG/AMP 2015 PS3: well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
PS4 requires that the variant be observed in affected individuals more often than in controls (case-control or case-enrichment evidence), or at minimum a documented set of affected probands. The literature pass for this case returned exactly two papers, both general ACMG/AMP framework documents (PMID:25741868, the 2015 standards; PMID:28492532, Sherloc), with candidate_criteria empty for both and no variant-specific content (variant_mentioned=false in paper_extracts.json; _variant_mentioned_count=0 in publications.json after screening for c.567T>C, p.(N189=), p.Asn189=, N189, 567T>C and NP_078918.3). No cohort, case series, or case-control study describing carriers of GALNT12 c.567T>C was identified, so there is no numerator (affected carriers) and no denominator from which an enrichment estimate could be formed. Numeric-threshold check as required for this group: no governing VCEP/gene-specific specification exists for GALNT12 (vcep_materials.json note 'no cspec url available'; final_classification_framework source='generic_acmg_fallback', cspec found=false in case_summary.json), and the generic ACMG/AMP 2015 framework defines no numeric odds-ratio or prevalence cutoff for PS4. No threshold value is present in the supplied published-calibration block for PS4 either, and none can be named from a publication, so no comparison operator and no substituted inequality can be constructed for this criterion. Under the generic framework PS4 remains a comparator-based criterion with no data on either side; this is an evidence gap rather than a negative result, so the criterion is not_assessed. Note for context only: the variant is catalogued in population reference samples (gnomAD v2.1 AC=207/AN=280,960, AF=0.000737; v4.1 AC=733/AN=1,611,402, AF=0.000455), which is control-side data and cannot substitute for affected-individual observations.
Generic ACMG/AMP 2015 PS4: variant prevalence is significantly increased in affected individuals compared with controls (case-control or case-enrichment evidence). No GALNT12-specific VCEP specification exists, so the generic rule governs.
PM3 requires that the variant be detected in trans with a pathogenic (or likely pathogenic) variant in a recessive disorder, i.e. a phase-resolved observation of a second, independently classified pathogenic allele on the opposite chromosome. No such data exist for NM_024642.5:c.567T>C in GALNT12. The case contains no proband genotype, pedigree, segregation, haplotype, or phasing information of any kind: all six ClinVar submissions for variation 416212 are germline clinical-testing submissions without an expert panel, none records a second GALNT12 allele or a phase relationship, and the only condition labels are 'Not Provided' and 'GALNT12-related condition'. No publication in the case corpus names c.567T>C or p.(Asn189=), so no trans observation can be sourced from the literature. The recessive gate itself also cannot be satisfied: the gene-level context artifact records an empty inheritance list for GALNT12 (no documented recessive disease association), and the NCBI gene summary describes GALNT12 as a dominant susceptibility locus for colorectal cancer. Because neither the variant-level trans observation nor a recessive disease context is available, the criterion cannot be resolved in either direction rather than being affirmatively met or refuted. No CSPEC or VCEP specification exists for GALNT12 (cspec found=false, framework_complete=false, source=generic_acmg_fallback), so the generic ACMG/AMP 2015 definition of PM3 applies. PM3 is therefore left as not_assessed and must not contribute to classification.
PM6 covers an assumed de novo variant whose parental identity has not been confirmed through testing; it is the weaker counterpart of PS2 and still presupposes that the variant was observed in a patient (Richards et al. 2015, PMID:25741868). For NM_024642.5:c.567T>C (GALNT12, NP_078918.3:p.(Asn189=)) there is no proband observation of the variant and no parental testing information of any kind, confirmed or merely assumed. The retrieved evidence is limited to variant-level database annotations (population frequency, ClinVar submissions, splice prediction, somatic COSMIC entry, hotspot lookup). Because even the minimal prerequisite - an observed de novo-appearing occurrence in a patient - is absent, PM6 cannot be applied; the gap is the missing proband and parental data, not conflicting evidence. No GALNT12 VCEP/CSPEC specification exists, so generic ACMG/AMP 2015 governs.
ACMG/AMP 2015 PM6 (Richards et al. 2015, PMID:25741868): assumed de novo variant with parental identity/testing unconfirmed
PP1 requires co-segregation of the variant with disease in multiple affected family members in a gene definitively known to cause the disease (Richards et al. 2015, PMID:25741868). For NM_024642.5:c.567T>C (GALNT12, NP_078918.3:p.(Asn189=)) no pedigree, no affected relatives, no genotyped family members, and therefore zero countable meioses are documented; co-segregation analysis cannot be performed. The retrieved GALNT12 literature characterises the gene as a moderate-penetrance colorectal cancer susceptibility gene rather than a gene definitively known to cause a high-penetrance Mendelian disorder, which would further constrain PP1 applicability even if family data existed. The criterion is not met on the available evidence and is recorded as not assessed because the pedigree data required to evaluate it are absent. No GALNT12 VCEP/CSPEC specification exists, so generic ACMG/AMP 2015 governs.
ACMG/AMP 2015 PP1 (Richards et al. 2015, PMID:25741868): co-segregation with disease in multiple affected family members
PP4 requires that the patient's phenotype (or family history) be highly specific for a disease with a single genetic aetiology, so that the match itself carries weight for the gene under evaluation. This case contains no proband-level clinical information at all: there are no HPO terms, no clinical description, no family history, and no referring phenotype anywhere in the case record or the retrieved literature. The only phenotype-adjacent data are ClinVar's condition labels attached to the exact variant record VCV000416212, which read 'Colorectal cancer, susceptibility to, 1', 'GALNT12-related disorder', 'not provided' and 'not specified' - these are record-level catalogue labels from submitting laboratories (several submissions carry no condition at all), not a description of this patient, and cannot establish phenotype specificity for an individual. Screening the two retrieved papers for the variant (terms c.567T>C, p.(N189=), p.Asn189=, N189, 567T>C, NP_078918.3) returned no carrier and therefore no phenotype to assess (variant_mentioned=false for both; publications.json _variant_mentioned_count=0). With no phenotype to compare against GALNT12's expected disease spectrum, PP4 is not evaluable and is recorded as not_assessed rather than not_met.
Generic ACMG/AMP 2015 PP4: the patient's phenotype or family history is highly specific for a disease with a single genetic aetiology. No GALNT12-specific VCEP specification exists, so the generic rule governs.
Not assessed. BS3 (Richards et al. 2015, PMID:25741868) requires well-established in vitro or in vivo functional studies showing no damaging effect on protein function. The variant is NM_024642.5:c.567T>C in GALNT12, a synonymous substitution encoding NP_078918.3:p.(Asn189=); no GALNT12 CSPEC or gene-specific VCEP specification exists, so the generic ACMG/AMP 2015 fallback applies (framework_complete=false). No functional assay of any kind is available for this variant in the retrieved sources: the source evidence, prefetch data, and literature pass contain no in vitro enzyme-activity, minigene, RT-PCR/splicing, protein-stability, or reporter result, and the only functional-themed paper retrieved (PMID:25741868) is the ACMG/AMP methodology paper, which does not mention GALNT12 or p.(N189=) (variant_mentioned=false). The available variant-level data are computational (SpliceAI max delta 0.017) and cannot substitute for a BS3 assay. Although ClinVar (ClinVarID 416212) carries Benign/Likely benign classifications from five laboratories, none supplies a variant-level functional experiment, and a classification from another source cannot be converted into BS3. With no qualifying benign functional result available, BS3 cannot be assessed from this dataset.
Generic ACMG/AMP 2015 BS3: well-established in vitro or in vivo functional studies showing a benign effect on the gene or gene product.
BS4 requires a demonstrated lack of segregation of the variant with disease among affected members of a family (Richards et al. 2015, PMID:25741868). For NM_024642.5:c.567T>C (GALNT12, NP_078918.3:p.(Asn189=)) no family members, affected or unaffected, were genotyped for this variant, so no non-segregation observation exists. Absence of pedigree data must not be converted into evidence of non-segregation, and the benign-direction argument cannot be constructed. The gene context is also relevant: GALNT12 is reported as a moderate-penetrance colorectal cancer susceptibility gene, so incomplete penetrance would weaken any segregation or non-segregation inference. The criterion is not met on the available evidence and is recorded as not assessed because the required family data are absent. No GALNT12 VCEP/CSPEC specification exists, so generic ACMG/AMP 2015 governs.
ACMG/AMP 2015 BS4 (Richards et al. 2015, PMID:25741868): lack of segregation in affected family members
BP2 requires a phase-resolved observation of this variant either in trans with a pathogenic variant in a fully penetrant dominant gene/disorder (for a benign-direction argument about a carrier individual) or in cis with a pathogenic variant under any inheritance pattern. Both branches require the variant to be observed alongside a second, pathogenic allele whose chromosomal phase relative to c.567T>C is known. No such observation exists in this case. There is no proband genotype or haplotype data: the six ClinVar submissions for variation 416212 report only classifications (4 laboratories Likely benign, 2 laboratories Benign) and condition labels ('Not Provided', 'GALNT12-related condition'), with no co-occurring variant, no cis/trans statement, and no family information; and no publication in the corpus mentions c.567T>C or p.(Asn189=), so no cis relationship with a pathogenic variant can be sourced from the literature. The variant's population record (gnomAD v2.1 total AF 0.00073676, 207/280960 alleles, 0 homozygotes; grpmc FAF 0.00117196, European non-Finnish AF 0.00134836) and its ClinVar benign-direction consensus are relevant to other criteria (frequency, PP5/BP6) but say nothing about cis/trans phase, and BP2 must not be inferred from them. Absent any phasing data the criterion can neither be met nor refuted, so it is recorded as not_assessed and contributes nothing to the classification. No CSPEC or VCEP specification exists for GALNT12, so the generic ACMG/AMP 2015 definition of BP2 governs.
BP5 applies when the variant is found in a patient who has an alternate molecular basis for the disease being investigated - for example a second, different pathogenic variant in another gene that already explains the phenotype, making this variant's contribution redundant. That reasoning requires proband-level knowledge: the affected individual's phenotype and the independent genetic finding that explains it. This case holds no such data. There is no proband record, no phenotype, no second-variant or multi-gene panel result, and no indication of what clinical question the testing was performed for. The ClinVar submissions for the exact variant provide no case-level information either: several carry empty condition labels and the remainder carry only generic ones ('Not Provided', 'GALNT12-related condition'), and the retrieved literature contains only the two general ACMG/AMP framework papers, neither of which mentions the variant or GALNT12 (variant_mentioned=false for both; publications.json _variant_mentioned_count=0 after screening c.567T>C, p.(N189=), p.Asn189=, N189, 567T>C, NP_078918.3). Without any documented case in which an alternate cause of disease was established, BP5 cannot be evaluated and is recorded as not_assessed. No numeric threshold applies (BP5 is not a threshold criterion, and no GALNT12 VCEP specification exists), so no comparison operator or inequality is required.
Generic ACMG/AMP 2015 BP5: the variant is found in a case with an alternate molecular basis for the disease (i.e. the phenotype is explained by another genetic finding). No GALNT12-specific VCEP specification exists, so the generic rule governs.
deterministic_prefilter: requires_null_variant_class
deterministic_prefilter: requires_missense
deterministic_prefilter: requires_missense
deterministic_prefilter: requires_stop_loss_or_nonrepeat_indel
deterministic_prefilter: requires_missense
deterministic_prefilter: requires_missense
deterministic_prefilter: requires_missense
deterministic_prefilter: requires_repeat_region_indel
Every interpreted variant in GALNT12, mapped onto the exon structure of NM_024642.5. Select a row to find it on the plot.
| Protein | Location | Classification | Link |
|---|
| European (Finnish)57 / 63,916 | 0.00089 |
| Ashkenazi Jewish21 / 29,586 | 0.00071 |
| Remaining individuals33 / 62,306 | 0.00053 |
| European (non-Finnish)618 / 1,179,622 | 0.000521 hom |
| Admixed American2 / 59,964 | 3.3e-05 |
| African/African American1 / 74,824 | 1.3e-05 |
| South Asian1 / 90,760 | 1.1e-05 |
| European (non-Finnish)173 / 128,304 | 0.0013 |
| Ashkenazi Jewish11 / 10,336 | 0.0011 |
| European (Finnish)18 / 24,764 | 0.00073 |
| Remaining individuals2 / 7,172 | 0.00028 |
| Admixed American2 / 35,318 | 5.7e-05 |
| South Asian1 / 30,330 | 3.3e-05 |
This variant is present in gnomAD v4.1 (AF= 0.000454883; MAF= 0.04549%, 733/1611402 alleles, homozygotes = 1) and has highest observed frequency in the European (Finnish) population (AF= 0.000891795; MAF= 0.08918%, 57/63916 alleles, homozygotes = 0); grpmax FAF= 0.00048901.
This variant is present in gnomAD v2.1 (AF= 0.00073676; MAF= 0.07368%, 207/280960 alleles, homozygotes = 0) and has highest observed frequency in the European (non-Finnish) population (AF= 0.00134836; MAF= 0.13484%, 173/128304 alleles, homozygotes = 0); grpmax FAF= 0.00117196.
This variant is present in gnomAD-Canada v1.0 (AF= 0.0005992590978426672, 11/18356 alleles, homozygotes = 0).
Scope first, identical determination as for PP3: NM_024642.5:c.567T>C is a synonymous substitution (NP_078918.3:p.(Asn189=)) in exon 3 of GALNT12, so the applicable BP4 path is the splice-prediction path (SpliceAI); there is no REVEL protein-impact path to use because a synonymous change has no amino-acid substitution to score, and the bundle's REVEL result is not found. Because BP4 is being taken from one path only, the missense-path score is neither used nor needed. SpliceAI Lookup for this exact variant gives a maximum delta score of 0.017 (DS_DG 0.017 is the largest of DS_AG 0.000 / DS_AL 0.007 / DS_DG 0.017 / DS_DL 0.007), with Pangolin splice-gain 0.01 and splice-loss -0.005. That maximum delta score is at or below the recommended SpliceAI no-impact cutoff of <= 0.1, which is the published BP4 (supporting) calibration for splice predictors (Jaganathan et al. 2019, Cell; PMID:30661751). Two independent directions agree that no splice consequence is predicted: no donor or acceptor is gained or strengthened (largest gain 0.017, at 21 bp downstream), and the acceptor-loss / donor-loss components are 0.000 and 0.007. The result is also internally consistent with the variant's position - c.567 sits 25 bp inside exon 3 (exon 3 spans c.542-731), i.e. away from the exon 3 acceptor (DP_AG -25) - and no cryptic splice site usage is predicted. BP4 is therefore met at supporting strength: multiple in-silico lines of splice evidence indicate no impact on splicing, which for a synonymous change is the relevant gene-product-level claim. This single splice prediction is credited here only; it is not also counted under BP7 (see BP7), and the framework's instruction that PP3/BP4 evidence come from one path only is respected. No VCEP/CSPEC specification for GALNT12 was available (cspec found=false; generic ACMG/AMP fallback, framework_complete=false), so the generic published threshold was applied verbatim.
Generic ACMG/AMP in-silico splice path: SpliceAI max delta <= 0.1 -> BP4 (supporting) (Jaganathan et al. 2019, Cell; PMID:30661751). Met: max delta 0.017.
Scope first. The variant is NM_024642.5:c.567T>C, which VariantValidator/Mutalyzer normalise to NP_078918.3:p.(Asn189=) (AAT>AAC, codon 189), i.e. a synonymous substitution in exon 3 of GALNT12 (chromosome 9q22.33; NC_000009.11:g.101589059T>C / NC_000009.12:g.98826777T>C) - not a missense change, and not a frameshift, nonsense or other protein-length-changing variant. PP3 for this variant therefore comes from the splice-prediction path (SpliceAI) only: there is no missense protein-impact claim to make for a synonymous change, and the REVEL protein-impact path is not applicable (the case bundle records REVEL as not found, and a missing missense score is not evidence and is not a reason to reach for a splice score in reverse). BayesDel is likewise not usable: no generic ACMG-strength calibration exists for it and no VCEP specification supplies one for GALNT12. No ClinGen CSPEC or gene-specific VCEP specification exists for GALNT12 (cspec found=false; final classification framework resolved to the generic ACMG/AMP fallback, framework_complete=false), so the generic ACMG/ClinGen-SVI calibration governs. SpliceAI Lookup for this exact variant reports DS_AG 0.000, DS_AL 0.007, DS_DG 0.017, DS_DL 0.007, giving a maximum delta score of 0.017 (the source's own summary sentence rounds this to 0.02); the nearest predicted cryptic acceptor is 25 bp upstream (DP_AG -25) and the nearest cryptic donor 21 bp downstream (DP_DG +21), i.e. neither is created or strengthened at the variant. Pangolin, reported alongside, gives splice-gain 0.01 and splice-loss -0.005. A maximum delta score of 0.017 is far below the recommended SpliceAI supporting cutoff for PP3 (max delta >= 0.2) and below the high-precision moderate cutoff (>= 0.5) (Jaganathan et al. 2019, Cell; PMID:30661751). No in-silico tool predicts a splice-altering effect, and no other computational evidence exists in this case that would support a damaging impact. PP3 is therefore not met. The variant's exonic position 25 bp into exon 3 (exon 3 spans c.542-731) is consistent with the absent splice signal, but PP3 rests on the predictor result, not on the location alone.
Generic ACMG/AMP in-silico splice path: SpliceAI max delta >= 0.2 -> PP3 (supporting); >= 0.5 -> PP3 (moderate) (Jaganathan et al. 2019, Cell; PMID:30661751). Not met: max delta 0.017.
Scope: BP7 applies only to a synonymous variant, and NM_024642.5:c.567T>C is synonymous (NP_078918.3:p.(Asn189=), AAT>AAC at codon 189), so BP7 is in scope and was evaluated rather than dismissed. BP7 as defined in the ACMG/AMP 2015 framework (Richards et al., PMID:25741868) requires a synonymous (silent) variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site, and for which the nucleotide is not highly conserved. The case contains exactly one item of in-silico evidence that speaks to the first requirement: the SpliceAI result for this variant - maximum delta score 0.017, with Pangolin splice-gain 0.01 and splice-loss -0.005 - which does predict no splice impact. That same single prediction, however, is the evidence already credited to BP4 at supporting strength for this variant (SpliceAI max delta <= 0.1). The ACMG/AMP framework states that predictions from different in-silico tools are treated as a single piece of evidence rather than independent pieces of evidence, and this framework additionally forbids double-counting one prediction across PP3/BP4/BP7; crediting the identical SpliceAI/Pangolin result under both BP4 and BP7 would count one prediction twice and inflate the benign evidence for this variant. The second BP7 element cannot be established either: no nucleotide-conservation metric for c.567 (phyloP, GERP, PhastCons or equivalent) is present anywhere in the case evidence, so the 'not highly conserved' clause of BP7 is unverified. BP7 is therefore not counted as an independent benign criterion for this variant; the prediction that would support it has already been applied once, under BP4.
BP7 (Richards et al. 2015, PMID:25741868): synonymous variant with no predicted impact on the splice consensus sequence or creation of a new splice site, and the nucleotide not highly conserved. Not credited here because the only supporting prediction is already counted under BP4 and the conservation element is unverified.
SpliceAI predicts no significant splice impact for this variant (max delta score = 0.02).
2 paper records listed. 1 cited in the classification. None of the listed paper records explicitly indicates that the variant is named.
A variant observed to have arisen de novo (parental samples test negative), is considered strong support for pathogenicity if the following conditions are met: Both parental samples were shown through identity testing to be the biological parents of the patient. Note: PM6 applies if identity is assumed but not confirmed.
Location Table 3 (Pathogenic criteria), PS2/PM6 'De novo variants' section
Context Consensus guideline document defining sequence-variant interpretation criteria and their evidence weights (PVS1, PS1-4, PM1-6, PP1-5; BA1, BS1-4, BP1-6); no experimental or case data generated.
The 2015 American College of Medical Genetics and Genomics-Association for Molecular Pathology (ACMG-AMP) guidelines were a major step toward establishing a common framework for variant classification. In practice, however, several aspects of the guidelines lack specificity, are subject to varied interpretations, or fail to capture relevant aspects of clinical molecular genetics.
Abstract (PubMed record); full text not retrieved
Screenshots represent source records captured during evidence collection. Select one to view the recorded source content.
The evidence statements recorded by the interpretation pipeline. They have not been reinterpreted by the page.
This variant is present in gnomAD v4.1 (AF= 0.000454883; MAF= 0.04549%, 733/1611402 alleles, homozygotes = 1) and has highest observed frequency in the European (Finnish) population (AF= 0.000891795; MAF= 0.08918%, 57/63916 alleles, homozygotes = 0); grpmax FAF= 0.00048901.
This variant is present in gnomAD v2.1 (AF= 0.00073676; MAF= 0.07368%, 207/280960 alleles, homozygotes = 0) and has highest observed frequency in the European (non-Finnish) population (AF= 0.00134836; MAF= 0.13484%, 173/128304 alleles, homozygotes = 0); grpmax FAF= 0.00117196.
This variant is present in gnomAD-Canada v1.0 (AF= 0.0005992590978426672, 11/18356 alleles, homozygotes = 0).
This variant has been reported in ClinVar as Likely benign (4 clinical laboratories) and as Benign (2 clinical laboratories). (ClinVarID = 416212)
SpliceAI predicts no significant splice impact for this variant (max delta score = 0.02).
This variant does not lie in a statistically significant hotspot. This variant has previously been reported in somatic cancers (COSMIC; COSV100899069, n = 1 times).
This variant does not lie in a statistically significant hotspot.