FGFR2 c.1516G>T, p.(Asp506Tyr)

Reasoning

PM2 is met at moderate strength (generic ACMG/AMP 2015): the variant is absent from gnomAD v2.1 and gnomAD-Canada v1.0, and in gnomAD v4.1 is observed at an extremely low total allele frequency of 0.00037% (6/1,614,106 alleles; grpmax FAF 1.772e-05) with a maximum subpopulation frequency of 0.00668% (East Asian), all well below the 0.1% non-VCEP PM2 threshold and consistent with extreme rarity expected for a pathogenic allele in a highly penetrant autosomal dominant FGFR2 craniosynostosis disorder.

Rule

generic ACMG/AMP 2015 PM2: absent or extremely low population AF (<0.1% non-VCEP gnomAD cutoff) in a highly penetrant dominant disorder

Evidence considered

• gnomAD v4.1: total AF 3.717e-06 (0.00037%), max subpopulation (East Asian) AF 6.682e-05 (0.00668%), grpmax FAF 1.772e-05, all below 0.1% PM2 threshold
• gnomAD v2.1 (GRCh37 10-123260385-C-A): absent
• gnomAD-Canada v1.0: absent
• Generic ACMG/AMP 2015 framework applied (no CSPEC/VCEP specification exists for FGFR2)

Sources

Reasoning

Under generic ACMG/AMP 2015 (Richards et al., PMID:25741868; applicable because no FGFR2 CSPEC/VCEP or local custom gene framework exists: cspec found=false, vcep_materials empty, gene_vcep_dir=null, framework_mode=generic_acmg, precedence_rank=3), PP3 is met when multiple lines of computational evidence support a deleterious effect on the gene or gene product. Two independent calibrated missense in-silico predictors both predict a deleterious effect for NM_000141.4(FGFR2):c.1516G>T (p.Asp506Tyr): REVEL score 0.667 falls in the ClinGen SVI-calibrated pathogenic band (>=0.644 supporting tier; below the 0.773 moderate tier) per the pipeline SVCv4 in-silico calibration table citing PMID 36413997 (verified), and BayesDel noAF score 0.100039 falls in the supporting-pathogenic tier (>=0.06, below the 0.13 moderate tier) per the same ladder (flagged placeholder/unverified), direction concordant with REVEL. SpliceAI predicts no splice impact (max delta 0.00; DS_AG=DS_AL=DS_DG=DS_DL=0.0), which is neutral for the missense path and does not contradict the deleterious predictions; the splice path simply does not contribute to PP3. Two concordant lines of calibrated computational evidence therefore support a deleterious effect, and PP3 is met at supporting strength (the generic ACMG/AMP 2015 framework caps PP3 at supporting; no gene-specific calibration exists to upgrade strength).

Rule

Generic ACMG/AMP 2015 PP3 (Richards et al., PMID:25741868; source_registry key generic_acmg_combination_rules): 'Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).' Supporting strength only; no FGFR2-specific calibration exists to modify strength.

Evidence considered

• REVEL score 0.667 (local lookup, revel-v1.3, chrom 10 pos 121500871 C>A) - ClinGen SVI-calibrated pathogenic supporting tier (>=0.644, verified against PMID 36413997); predicts deleterious missense effect
• BayesDel noAF score 0.100039 (local lookup, BayesDel_170824_noAF_chr10) - supporting-pathogenic tier (>=0.06) in the pipeline SVCv4 ladder (placeholder calibration); predicts deleterious missense effect, concordant with REVEL
• SpliceAI max delta score 0.00 (DS_AG=DS_AL=DS_DG=DS_DL=0.0, variant_consequence=sequence_variant) - no predicted splice impact; neutral for PP3, does not contradict the missense deleterious predictions
• Generic ACMG/AMP 2015 PP3 definition (PMID:25741868): 'Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.)'

Still missing

• BayesDel noAF calibration is flagged placeholder/unverified in the pipeline SVCv4 in-silico threshold table (insilico.py); it is used only as a corroborating line, with REVEL (verified, PMID 36413997) as the primary predictor
• aGVGD score not retrieved for this case; PP3 does not depend on it
• No conservation-based (e.g., PhyloP/GERP) or additional splicing-predictor lines were available; the two concordant calibrated missense predictors are sufficient for the supporting-strength determination
• HCI prior not available (FGFR2 not supported in the HCI prior model); not required for PP3

Sources

Reasoning

PS1 (Richards et al. 2015, PMID:25741868) requires a previously established pathogenic variant producing the same amino acid change as this variant, regardless of nucleotide change. For FGFR2 NP_000132.3:p.(Asp506Tyr), deterministic codon analysis shows the only coding change capable of producing Tyr at residue 506 is c.1516G>T itself (Asp codons GAT/GAC -> Tyr codons TAT/TAC require G>T at the first codon position; no alternate nucleotide change at this codon yields Tyr). ClinVar was searched with six term sets (c.1516G>T, NM_000141.4:c.1516G>T, both genomic HGVS forms) and returned no record ('This variant is absent from ClinVar'), the literature pass returned 0 PMIDs, and the pm5_candidates artifact collected 0 comparator candidates. There is therefore no established pathogenic variant with the same amino acid change anywhere in the consulted evidence, and PS1 cannot be met.

Rule

generic ACMG/AMP 2015 PS1 (PMID:25741868): same amino acid change as a previously established pathogenic variant regardless of nucleotide change

Evidence considered

• ClinVar search (FGFR2[GENE] AND c.1516G>T[VARNAME], NM_000141.4:c.1516G>T, NC_000010.10:g.123260385C>A and NC_000010.11:g.121500871C>A): no result - variant absent
• literature_pass.json: total_pmids=0; no publication mentions c.1516G>T or p.Asp506Tyr
• Deterministic codon analysis: p.Asp506Tyr is reachable only via c.1516G>T (GAT/TAC), so 'same amino acid change' is equivalent to this exact variant

Still missing

• No residue-level ClinVar scan for alternate transcripts/aliases of FGFR2 p.Asp506Tyr was performed, though the unique-codon argument makes an alternate-nucleotide same-change variant impossible at this position.

Sources

Reasoning

PM1 (Richards et al. 2015, PMID:25741868) requires the variant to be in a mutational hotspot or a critical and well-established functional domain without benign variation. The dedicated hotspot source (CancerHotspots, evidence.json results.hotspots) returned no statistically significant hotspot for FGFR2 D506 ('This variant does not lie in a statistically significant hotspot'), directly failing the mutational-hotspot arm. The critical-functional-domain arm could not be evaluated: the case bundle contains no authoritative domain-boundary annotation for residue 506 (UniProt/Pfam kinase-domain mapping is not present) and therefore no evidence that the surrounding domain harbors no benign variation. FGFR2 is a receptor tyrosine kinase (gene_context) and the variant falls in exon 11, which is consistent with the kinase-encoding region, but this is not sufficient to establish a 'critical and well-established functional domain with no benign variation' under generic PM1. PM1 is not met on the available evidence; the domain arm requires human review with an external domain/constraint resource.

Rule

generic ACMG/AMP 2015 PM1 (PMID:25741868): located in a mutational hotspot and/or critical and well-established functional domain without benign variation

Evidence considered

• CANCERHOTSPOTS (evidence.json results.hotspots, query 'FGFR2 D506'): no hotspot row found - not a statistically significant hotspot
• OncoKB (evidence.json results.oncokb): no variant-specific functional or hotspot context for D506Y (Unknown Oncogenic Effect, not reviewed)
• Prefetch/VariantValidator: variant is in exon 11 of NM_000141.4; no domain boundary annotation available in bundle

Still missing

• Authoritative domain annotation for residue 506 (e.g., UniProt/Pfam kinase domain boundaries) not present in the case bundle
• Benign-variation-within-domain assessment (e.g., gnomAD missense constraint within the domain) not possible from bundle data

Reasoning

PP5 is policy-restricted in this group to an exact-variant ClinVar expert-panel classification: a Pathogenic/Likely pathogenic classification by an expert panel (3-star review status) supports PP5; ordinary laboratory submissions, aggregate ClinVar labels, and non-expert assertions never trigger it. FGFR2 c.1516G>T is entirely absent from ClinVar (evidence.json results.clinvar search_status=no_result; case_summary.json compact_evidence.clinvar: variation_ids=[], submissions=[], review_status=''). No exact-variant expert-panel classification of any kind exists, so the sole trigger for PP5 is definitively absent and the criterion is not met.

Rule

Policy-restricted PP5: exact-variant ClinVar expert-panel (3-star) Pathogenic/Likely pathogenic classification only; laboratory submissions and aggregate labels never trigger

Evidence considered

• ClinVar exact-variant search returned no result (evidence.json results.clinvar)
• case_summary.json compact_evidence.clinvar: variation_ids empty, submissions empty

Still missing

• No ClinVar expert-panel classification exists for this exact variant because the variant is absent from ClinVar; PP5 would require a future expert-panel Pathogenic/Likely pathogenic submission for this exact variant.

Sources

Reasoning

BA1 is not met: the highest observed population frequency of NM_000141.4:c.1516G>T (p.Asp506Tyr) in FGFR2 is 0.00668% in the East Asian gnomAD v4.1 subpopulation (3/44,900 alleles), and the total gnomAD v4.1 allele frequency is 0.00037% (6/1,614,106 alleles), both orders of magnitude below the 1% non-VCEP BA1 threshold; the variant is absent from gnomAD v2.1 and gnomAD-Canada v1.0, and no threshold artifact (e.g., population-specific founder enrichment) approaches the cutoff.

Rule

generic ACMG/AMP 2015 BA1: population AF >1% (non-VCEP gnomAD cutoff)

Evidence considered

• gnomAD v4.1: total AF 3.717e-06 (0.00037%, 6/1,614,106 alleles), 0 homozygotes; best subpopulation East Asian AF 6.682e-05 (0.00668%, 3/44,900 alleles), 0 homozygotes; grpmax FAF 1.772e-05
• gnomAD v2.1 (GRCh37 10-123260385-C-A): absent
• gnomAD-Canada v1.0: absent

Sources

Reasoning

BS1 is not met: the highest subpopulation allele frequency (East Asian 0.00668%) is more than 40-fold below the 0.3% non-VCEP BS1 threshold, and the total gnomAD v4.1 frequency (0.00037%) is even lower; such a frequency is far below what would be expected for a benign variant in a highly penetrant autosomal dominant FGFR2 craniosynostosis disorder, whose pathogenic variants are typically absent or ultra-rare in population cohorts.

Rule

generic ACMG/AMP 2015 BS1: population AF >0.3% (non-VCEP gnomAD cutoff)

Evidence considered

• gnomAD v4.1: total AF 3.717e-06 (0.00037%); best subpopulation (East Asian) AF 6.682e-05 (0.00668%), well below 0.3% BS1 threshold
• gnomAD v2.1: absent
• gnomAD-Canada v1.0: absent

Still missing

• No quantitative FGFR2 craniosynostosis disease-prevalence estimate was available in the case bundle to compute an expected pathogenic allele frequency; BS1 was judged against the fixed 0.3% non-VCEP cutoff, which the observed frequencies do not approach.

Sources

Reasoning

BS2 is not met: no homozygous carriers of NM_000141.4:c.1516G>T are observed in gnomAD v4.1 (0 homozygotes across 1,614,106 alleles; only 6 heterozygous alleles total), and the variant is absent from ClinVar, so there is no observation of the variant in homozygous state or in healthy adult individuals with full-penetrance-dominant FGFR2 craniosynostosis phenotypes as required by BS2.

Rule

generic ACMG/AMP 2015 BS2: variant observed in healthy adult (homozygous, or heterozygous for dominant disorder) with full penetrance expected at early age

Evidence considered

• gnomAD v4.1: 0 homozygotes; total AC 6 / AN 1,614,106, all heterozygous
• gnomAD v2.1: absent
• ClinVar: no submissions, this variant is absent from ClinVar

Still missing

• No dedicated healthy-adult carrier cohort data beyond gnomAD (e.g., ClinVar submissions with healthy-individual observations) are available; variant is absent from ClinVar entirely.

Sources

Reasoning

BP1 (Richards et al. 2015, PMID:25741868) applies as a supporting benign criterion only when the gene's disease mechanism is predominantly loss-of-function/truncating, making a missense variant unlikely to be pathogenic. The opposite is true for FGFR2: germline FGFR2 disease (Apert, Crouzon, Pfeiffer, Jackson-Weiss, Beare-Stevenson and related craniosynostosis syndromes) is established to be caused by heterozygous activating missense variants, per the gene mechanism literature bundled under pvs1_gene_context (PMID:38265560: 'Heterozygous activating non-mosaic germline variants in FGFR2 have been linked to numerous autosomal dominantly inherited disorders...'; PMID:41572294: 'Dominant missense mutations in FGFR1, FGFR2, and FGFR3 have been associated with at least 20 congenital disorders characterized by skeletal abnormalities') and the gene_context summary ('Germline mutations in FGFR2 cause several inherited craniosynostosis syndromes'). Missense is a major, not minor, mechanism of FGFR2 germline disease, so the BP1 premise (primarily truncating variants cause disease) is contradicted and BP1 is not met. Note: this conclusion is independent of the PVS1-group's lof_mechanism_supported finding, which concerns a different (null-variant) question.

Rule

generic ACMG/AMP 2015 BP1 (PMID:25741868): missense variant in a gene for which primarily truncating variants are known to cause disease

Evidence considered

• pvs1_gene_context literature (PMID:38265560, PMID:41572294, PMID:34829964, PMID:40362441): FGFR2 germline disease is activating-missense-driven, not truncating-only
• gene_context_FGFR2.json gene_summary: FGFR2 germline mutations cause craniosynostosis syndromes; receptor tyrosine kinase mechanism

Sources

Reasoning

BP2 (Richards et al. 2015) requires observation of the variant in trans with a pathogenic variant for a fully dominant disease/gene, or in cis with a pathogenic variant in any inheritance pattern. FGFR2 germline disorders are fully dominant (see PM3 rationale), so the criterion is applicable in principle, but there is no evidence to satisfy it: the variant is absent from ClinVar (search_status 'no_result', zero submissions), literature_pass.json reports 0 PMIDs, and the publications/ directory and fulltext_map.json are empty. The only occurrences of the variant are population-allele counts in gnomAD v4.1 (6/1,614,106 alleles, AF 3.71723e-06, 0 homozygotes; absent from gnomAD v2.1 and gnomAD-Canada), which carry no proband phenotype, phasing, or co-occurrence information. No proband, family segregation, or cis/trans phase data exist in this case bundle, so BP2 is not met.

Rule

ACMG/AMP 2015 (PMID 25741868): Observed in trans with a pathogenic variant for a fully dominant disease/gene, OR observed in cis with a pathogenic variant in any inheritance pattern (benign supporting).

Evidence considered

• generic_acmg_combination_rules
• final_classification_framework
• clinvar
• gnomad_v4

Still missing

• No clinical or family observations of this variant exist (ClinVar absent, 0 literature PMIDs), so no phase-resolved cis/trans co-occurrence with a pathogenic variant can be established.
• If proband or family phase/segregation data become available, BP2 must be re-evaluated.

Sources

Reasoning

Under generic ACMG/AMP 2015 (PMID:25741868), BP4 requires multiple lines of computational evidence suggesting no impact on the gene or gene product. The only benign-direction line available is SpliceAI (max delta 0.00, no predicted splice impact for this exonic missense, c.1516G>T located in exon 11 of NM_000141.4, 76 nt from exon start and 45 nt from exon end); this is a single splice-path line and does not address protein function. The two calibrated missense predictors directly contradict BP4: REVEL 0.667 (ClinGen SVI pathogenic supporting band) and BayesDel noAF 0.100039 (supporting-pathogenic tier) both predict a deleterious effect on FGFR2. The computational evidence as a whole is concordant for a deleterious effect, not for no impact, so BP4 is not met. The SpliceAI non-effect is not double-counted as BP7 (variant is missense, not synonymous) and not weighed against PP3 (PP3 rests on the missense predictions, not on SpliceAI).

Rule

Generic ACMG/AMP 2015 BP4 (Richards et al., PMID:25741868; source_registry key generic_acmg_combination_rules): 'Multiple lines of computational evidence suggest no impact on the gene or gene product (conservation, evolutionary, splicing impact, etc.).' Requires concordant benign-direction computational evidence; discordant missense predictions preclude application.

Evidence considered

• SpliceAI max delta score 0.00 - single benign-direction line (splice path only): no predicted splice impact for c.1516G>T (exonic, exon 11)
• REVEL score 0.667 - pathogenic supporting band; contradicts BP4
• BayesDel noAF score 0.100039 - supporting-pathogenic tier; contradicts BP4
• Generic ACMG/AMP 2015 BP4 definition (PMID:25741868): 'Multiple lines of computational evidence suggest no impact on the gene or gene product (conservation, evolutionary, splicing impact, etc.)'

Still missing

• No additional benign-direction computational lines available (e.g., benign-calibrated missense predictors or conservation data showing no impact); not required for the not_met determination given the discordant missense predictions
• No RNA-based experimental confirmation of absent splice impact; SpliceAI 0.00 is in silico only and was not used as BP4 evidence on its own

Sources

Reasoning

BP6 is policy-restricted in this group to an exact-variant ClinVar expert-panel classification: a Benign/Likely benign classification by an expert panel (3-star review status) supports BP6; ordinary laboratory submissions, aggregate ClinVar labels, and non-expert assertions never trigger it. FGFR2 c.1516G>T is entirely absent from ClinVar (evidence.json results.clinvar search_status=no_result; case_summary.json compact_evidence.clinvar: variation_ids=[], submissions=[], review_status=''). No exact-variant expert-panel benign classification exists, so the sole trigger for BP6 is definitively absent and the criterion is not met.

Rule

Policy-restricted BP6: exact-variant ClinVar expert-panel (3-star) Benign/Likely benign classification only; laboratory submissions and aggregate labels never trigger

Evidence considered

• ClinVar exact-variant search returned no result (evidence.json results.clinvar)
• case_summary.json compact_evidence.clinvar: variation_ids empty, submissions empty

Still missing

• No ClinVar expert-panel classification exists for this exact variant because the variant is absent from ClinVar; BP6 would require a future expert-panel Benign/Likely benign submission for this exact variant.

Sources

Reasoning

PS2 requires a de novo occurrence of the variant in the proband with both maternity and paternity confirmed (or an equivalent molecular confirmation of parentage). The case bundle contains no proband-level clinical or sequencing data, no parental testing results, and no parentage confirmation for FGFR2 c.1516G>T (p.Asp506Tyr). The variant is absent from ClinVar (clinvar.search_status = no_result), literature_pass reports 0 PMIDs, and fulltext_map.json is empty, so no published de novo report of this variant exists to evaluate. There is no evidence on which to base a PS2 call in either direction.

Rule

PS2 (Richards et al. 2015, PMID:25741868): De novo variant (both maternity and paternity confirmed) in a gene definitively known to cause the disease

Still missing

• Proband phenotype and genotype data for FGFR2 c.1516G>T
• Parental testing results for both biological parents
• Parentage confirmation (maternity and paternity) or documented alternative confirmation method
• Any published de novo case report of c.1516G>T (p.Asp506Tyr)

Reasoning

PS3 requires well-established functional studies demonstrating a damaging effect on gene/protein function (Richards et al. 2015, PMID:25741868). No functional assay evidence exists for FGFR2 c.1516G>T (p.Asp506Tyr) anywhere in the case: zero PMIDs were retrieved across all evidence sources (literature_pass total_pmids=0, fulltext_map empty), ClinVar/COSMIC/Cancer Hotspots have no entry for this variant, OncoKB explicitly reports no variant-specific reviewed functional evidence ('Unknown Oncogenic Effect', variant_specific=false), and the available in silico predictors (SpliceAI max delta 0.00, REVEL 0.667, BayesDel 0.100039) are not functional assays and cannot satisfy PS3 under generic ACMG/AMP. Evidence is insufficient to assess; criterion not met on the available record.

Rule

generic_acmg_fallback (Richards et al. 2015, PMID:25741868): PS3 = well-established functional studies showing a damaging effect on gene/protein function; in silico predictions do not qualify as functional evidence

Evidence considered

• OncoKB FGFR2 D506Y page: no variant-specific reviewed functional evidence identified; classification 'Unknown Oncogenic Effect' (not_reviewed=true)
• SpliceAI max delta score 0.00: in silico splice prediction only, no splice impact predicted
• REVEL score 0.667: in silico missense pathogenicity predictor, not a functional assay
• BayesDel noAF score 0.100039: in silico missense predictor, not a functional assay
• ClinVar absent; COSMIC absent; Cancer Hotspots: no statistically significant hotspot

Still missing

• No validated variant-specific functional assay for p.(Asp506Tyr) retrieved from any case source (e.g., FGFR2 kinase activity/autophosphorylation, downstream MAPK/PLC-gamma signaling, cell proliferation/transformation, or receptor surface-expression assays)
• No published literature identified (0 PMIDs across all evidence sources) to screen for variant-specific functional studies

Sources

Reasoning

PS4 (ACMG/AMP 2015, PMID:25741868) requires case-control or case-series evidence showing the prevalence of the variant is significantly increased in affected individuals compared with controls. No such data exist for FGFR2 c.1516G>T: the variant is absent from ClinVar (evidence.json results.clinvar search_status=no_result), literature_pass.json reports total_pmids=0, the publications directory is empty, and fulltext_map.json is empty. With no cohort, case-series, or enrichment data available, PS4 cannot be evaluated; the evidence is insufficient rather than contradictory.

Rule

ACMG/AMP 2015 PS4: prevalence of variant in affected individuals significantly increased versus controls (case-control studies)

Evidence considered

• ClinVar exact-variant search returned no result (evidence.json results.clinvar; case_summary.json compact_evidence.clinvar)
• literature_pass.json total_pmids=0; publications directory empty; fulltext_map.json empty

Still missing

• No case-control or case-series studies reporting FGFR2 c.1516G>T were identified; a targeted literature/cohort search for variant-specific enrichment data is required before PS4 can be scored.

Reasoning

PM5 (Richards et al. 2015, PMID:25741868) requires a novel missense change at a residue where a DIFFERENT pathogenic missense variant has been previously established. The pipeline's PM5 candidate harvesting was not executed: pm5_candidates.json reports found=false, pm5_mode='unknown', 'Unable to confirm classic same-residue PM5 semantics safely; skipping automatic candidate harvesting', with 0 total candidates seen and an empty ClinVar query strategy (query_strategy.clinvar_queries: []). The only ClinVar search performed was for the exact variant c.1516G>T (absent), which would not capture a different pathogenic missense at residue 506 (e.g., p.Asp506Asn, p.Asp506His, p.Asp506Gly). No same-residue comparator is mentioned in any bundle source and literature_pass returned 0 PMIDs. Because the comparator space for PM5 (established pathogenic missense at Asp506 other than p.Asp506Tyr) was never systematically searched, the criterion cannot be affirmed or confidently excluded; per the artifact's own instruction, PM5 must not be applied from this artifact alone. A manual residue-level ClinVar/PubMed scan for pathogenic FGFR2 missense at residue 506 is required.

Rule

generic ACMG/AMP 2015 PM5 (PMID:25741868): novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before

Evidence considered

• pm5_candidates.json: found=false, pm5_mode='unknown', eligible_for_classic_pm5_search=false, total_candidates_seen=0, same_residue_candidates=0
• ClinVar: no entry for c.1516G>T; no same-residue comparator identified in any consulted source
• literature_pass.json: total_pmids=0; no comparator literature available

Still missing

• No residue-level search for established pathogenic missense variants at FGFR2 residue 506 (different amino acid) was performed; pm5_candidates.json skipped candidate harvesting (pm5_mode=unknown)
• ClinVar query strategy executed no comparator queries; manual ClinVar/PubMed residue-506 review required

Sources

Reasoning

PM6 applies to an assumed de novo variant where maternity and paternity are not confirmed. Like PS2, this requires proband-level data showing the variant arose de novo; no such data exists for FGFR2 c.1516G>T (p.Asp506Tyr). The case bundle has no proband or parental sequencing results, the variant is absent from ClinVar, and no literature PMIDs were retrieved (literature_pass.total_pmids = 0). Without any observation of a de novo event, PM6 cannot be assessed.

Rule

PM6 (Richards et al. 2015, PMID:25741868): Assumed de novo, but without confirmation of paternity and maternity

Still missing

• Proband genotype demonstrating the variant is not inherited
• Parental testing results (present but without parentage confirmation)
• Any published case reporting the variant as de novo without confirmed parentage

Reasoning

PP1 requires co-segregation of the variant with disease in multiple affected family members in a gene definitively known to cause the disease. FGFR2 is definitively associated with autosomal dominant craniosynostosis/skeletal syndromes, so the criterion is potentially applicable, but the case bundle contains no family or pedigree data whatsoever: no affected relatives tested, no segregation counts, no meioses reported. ClinVar has no entry for this variant and no publications were retrieved, so no segregation evidence exists to evaluate. PP1 cannot be met or refuted on current evidence.

Rule

PP1 (Richards et al. 2015, PMID:25741868): Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease

Still missing

• Pedigree/family history for the proband
• Genotype-phenotype segregation data in affected family members (number of informative meioses)
• Any published family segregating c.1516G>T (p.Asp506Tyr)

Reasoning

PP2 (Richards et al. 2015, PMID:25741868) is a conjunctive criterion requiring BOTH (a) a low rate of benign missense variation in the gene and (b) missense being a common mechanism of disease. Arm (b) is supported: the gene-mechanism literature bundled under pvs1_gene_context shows FGFR2 germline disease is dominated by heterozygous activating missense variants causing craniosynostosis syndromes (e.g., PMID:38265560: 'Heterozygous activating non-mosaic germline variants in FGFR2 have been linked to numerous autosomal dominantly inherited disorders including several craniosynostoses and skeletal dysplasia syndromes'; PMID:41572294: 'Dominant missense mutations in FGFR1, FGFR2, and FGFR3 have been associated with at least 20 congenital disorders'). Arm (a) cannot be evaluated: the bundle contains no quantitative missense-constraint metric (no gnomAD missense Z-score / o/e ratio, no benign-missense-vs-pathogenic-missense ratio in ClinVar), and the HCI prior is unavailable for FGFR2 (hci_prior: gene_not_supported, custom_pp2_score null). Because one required arm lacks any data, PP2 is not assessable from this case bundle; a human reviewer must supply gene-level missense constraint and benign-missense data to complete the adjudication.

Rule

generic ACMG/AMP 2015 PP2 (PMID:25741868): missense variant in a gene that has a low rate of benign missense variation and in which missense variants are a common mechanism of disease

Evidence considered

• pvs1_gene_context literature (PMID:38265560, PMID:41572294, PMID:34829964, PMID:40362441, PMID:41062690): FGFR2 germline disease mechanism is heterozygous activating missense (craniosynostosis syndromes) - supports arm (b)
• case_summary hci_prior: found=false (gene_not_supported); no missense constraint data available

Still missing

• No gnomAD missense constraint metric (Z-score / observed-expected ratio) or ClinVar benign-missense ratio for FGFR2 in the bundle
• HCI prior / custom PP2 score unavailable (hci_prior: gene_not_supported, custom_pp2_score: null)

Sources

Reasoning

PP4 (ACMG/AMP 2015, PMID:25741868) requires a phenotype or family history highly specific for a disease with a single genetic etiology. The case bundle contains no proband phenotype, clinical description, or family history section (case_summary.json contains only variant-annotation data: ClinVar, gnomAD, SpliceAI, REVEL, BayesDel, OncoKB, COSMIC, hotspots, PVS1 context). Without any phenotype information, PP4 cannot be assessed. Gene-level FGFR2 germline disease context (craniosynostosis syndromes) exists in pvs1_gene_context.json but is not proband-specific and none of its sources mention this exact variant, so it cannot substitute for PP4 phenotype evidence.

Rule

ACMG/AMP 2015 PP4: phenotype or family history highly specific for a disease with a single genetic etiology

Evidence considered

• case_summary.json contains no phenotype/family-history fields
• literature_pass.json total_pmids=0; no publications available to supply phenotype data

Still missing

• Proband phenotype and family history data are absent from the case bundle; a clinical phenotype description is required to evaluate phenotype specificity for an FGFR2-related disorder.

Reasoning

BS3 requires well-established functional studies demonstrating no damaging effect on gene/protein function (Richards et al. 2015, PMID:25741868). No such evidence exists for FGFR2 c.1516G>T (p.Asp506Tyr): no functional assay data were retrieved from any source, OncoKB classifies the variant as 'Unknown Oncogenic Effect' (an annotation of unknown effect, not an assay demonstrating benign function), and in silico predictions (SpliceAI max delta 0.00; REVEL 0.667; BayesDel 0.100039) do not constitute functional evidence under generic ACMG/AMP. Evidence is insufficient to assess; criterion not met on the available record.

Rule

generic_acmg_fallback (Richards et al. 2015, PMID:25741868): BS3 = well-established functional studies showing no damaging effect on gene/protein function; in silico predictions do not qualify as functional evidence

Evidence considered

• OncoKB FGFR2 D506Y page: 'Unknown Oncogenic Effect' with no variant-specific reviewed functional evidence; not an assay of benign function
• SpliceAI max delta score 0.00: in silico splice prediction only; addresses splicing, not protein function
• REVEL score 0.667: in silico missense pathogenicity predictor, not a functional assay
• BayesDel noAF score 0.100039: in silico missense predictor, not a functional assay
• ClinVar absent; COSMIC absent; Cancer Hotspots: no statistically significant hotspot

Still missing

• No functional assay demonstrating normal/no-damaging-effect FGFR2 function for p.(Asp506Tyr)
• No published literature identified (0 PMIDs across all evidence sources) to screen for variant-specific functional studies

Sources

Reasoning

BS4 requires observed non-segregation of the variant with disease in affected family members (an affected family member who does not carry the variant, or an unaffected carrier with the variant). No family testing data exists for FGFR2 c.1516G>T (p.Asp506Tyr) in this case: there are no pedigrees, no affected relatives tested, and no literature reports (ClinVar absent, literature_pass.total_pmids = 0). The absence of any reported non-segregation is not evidence of segregation, so BS4 is not met; it cannot be assessed without family data. Population frequency data (gnomAD v4.1 AF 3.7e-6) is not BS4 evidence and was not used here.

Rule

BS4 (Richards et al. 2015, PMID:25741868): Non-segregation with disease in affected family members

Still missing

• Genotype data for affected family members to test for non-segregation
• Any published report of non-segregation of c.1516G>T (p.Asp506Tyr) with FGFR2-related disease

Reasoning

BP5 (ACMG/AMP 2015, PMID:25741868) applies when the variant is found in a case with an alternate molecular basis for disease, indicating it is likely not causal. The case bundle contains no proband-level data whatsoever: no clinical case report, no segregation information, no alternative-diagnosis workup, and no publications (literature_pass total_pmids=0). Whether the carrier has an alternate molecular cause cannot be determined, so BP5 cannot be assessed.

Rule

ACMG/AMP 2015 BP5: variant found in a case with an alternate molecular basis for disease

Evidence considered

• No proband-level clinical data present in case_summary.json or evidence.json
• literature_pass.json total_pmids=0; no publications available

Still missing

• No proband-level data on alternative molecular diagnoses or second genetic findings; a clinical case description or study reporting this variant in a patient with another established cause is required.