ROS1 c.6068A>T, p.(Tyr2023Phe)

Reasoning

Generic ACMG/AMP 2015 (PMID:25741868) defines PM2 as the variant being absent in controls / population databases (ExAC/gnomAD and similar), i.e., extremely rare in the general population. NM_002944.2:c.6068A>T is absent (AF = 0) from gnomAD v2.1 exomes, gnomAD v4.1 exomes, and gnomAD-Canada v1.0 genomes - three independent population datasets spanning GRCh37 and GRCh38, exome and genome data. AF = 0 satisfies the PM2 absence criterion and is well below any operational rarity threshold for support. PM2 is met at supporting strength. Confidence is moderate rather than high because the bundle records only variant-page 'absent' status with empty source_data (no per-position allele counts, sample sizes, or coverage metrics), so per-position coverage quality in this large gene (ROS1, 6q22.1, variant in exon 38 near the 3' end of the CDS) could not be independently verified from the evidence bundle.

Rule

generic_acmg_2015:PM2_absent_in_population_databases

Evidence considered

• gnomAD v2.1 variant page 6-117638373-T-A: absent (search_status 'absent')
• gnomAD v4.1 variant page 6-117317210-T-A: absent (search_status 'absent')
• gnomAD-Canada v1.0 (HostSeq genomes) API lookup 6-117317210-T-A: absent
• PM2 definition (absent in population databases/controls) per Richards et al. 2015 (PMID:25741868); absence across gnomAD v2.1 + v4.1 + gnomAD-Canada supports supporting-strength PM2

Still missing

• gnomAD source_data in the bundle is empty: per-population allele counts, total allele numbers, and per-position coverage/depth for this locus were not captured; a reviewer could confirm against the gnomAD variant pages or VCF to rule out a sequencing/coverage artifact as the cause of absence.
• No homozygote/allele-count table was available to quantify how many alleles were surveyed at this position in each dataset.

Sources

Reasoning

PM1 requires the variant to lie in a mutational hotspot and/or a critical, well-established functional domain without benign variation. The case-bundle hotspots check (cancerhotspots.org query 'ROS1 Y2023') returned no statistically significant hotspot. No ROS1-specific critical functional domain map is present in vcep_materials.json (no CSPEC/VCEP, no curated domain payload), so no curated domain-based PM1 support exists in this case. p.Tyr2023 falls in the C-terminal region (exon 38) of this receptor tyrosine kinase, within the broadly annotated tyrosine kinase region by general protein-domain annotations, but that annotation is not backed by a named, curated source in this case bundle and generic-ACMG PM1 is not applied to whole domains as a default. PM1 is therefore not met on available evidence; flagged for human review in case a named ROS1 domain/hotspot map becomes available.

Rule

ACMG/AMP 2015 PM1: Located in a mutational hot spot and/or critical and well-established functional domain (e.g., active site of an enzyme) without benign variation (Richards et al. 2015, PMID 25741868; generic_acmg_combination_rules).

Evidence considered

• Case-bundle hotspots evidence (cancerhotspots.org, query 'ROS1 Y2023'): no statistically significant hotspot found (evidence.json hotspots result, search_status no_result).
• vcep_materials.json: no CSPEC url, no files, no curated domain guidance; final_classification_framework not found (generic ACMG fallback).
• variant annotation (prefetch.json variant_validator): variant in exon 38, protein NP_002935.2 p.(Tyr2023Phe), C-terminal region of ROS1.

Still missing

• No ROS1-specific critical functional domain map (VCEP/curated, named source) was available to confirm or refute kinase-domain membership; hotspot query was negative. Confirm critical-domain status and low benign-variation rate at this residue from a named publication before applying PM1.

Reasoning

No available calibrated computational line supports a deleterious effect. SpliceAI max delta score is 0.01 (DS_AG=0.01, DS_AL=0.00, DS_DG=0.00, DS_DL=0.00), far below the published splice-altering thresholds of 0.20 (high recall) and 0.50 (high precision) from Jaganathan et al., Cell 2019 (PMID 30661751), so the splice-impact path provides no PP3 support. REVEL score 0.52 falls in the uninformative interval of the ClinGen SVI calibration (Pejaver et al., Am J Hum Genet 2022, PMID 36413997): it is below the PP3 supporting threshold (REVEL >= 0.773) and above the BP4 supporting threshold (REVEL <= 0.016), so it supports neither PP3 nor BP4. BayesDel noAF score -0.119347 is treated as insufficiently calibrated (no verified published threshold available to this pipeline, per group instructions) and is not counted. aGVGD was not available. No gene-specific VCEP lookup spreadsheet exists for ROS1 to pre-assign this criterion. With zero predictor lines in the deleterious direction, PP3 is not met.

Rule

Generic ACMG/AMP 2015 (Richards et al., PMID 25741868; source_registry key 'generic_acmg_combination_rules'): PP3 - multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.). No ROS1 CSPEC/VCEP or gene-specific computational lookup spreadsheet exists (framework_mode=generic_acmg, precedence_rank=3).

Evidence considered

• SpliceAI max delta score 0.01 (DS_AG=0.01, DS_AL=0.00, DS_DG=0.00, DS_DL=0.00; pangolin SG=0.0, SL=-0.02) predicts no splice impact; below the published splice-altering thresholds of 0.20 (high recall) and 0.50 (high precision) reported by Jaganathan et al., Cell 176:535-548, 2019 (PMID 30661751).
• REVEL score 0.52 is below the ClinGen SVI PP3 supporting threshold (REVEL >= 0.773) and above the BP4 supporting threshold (REVEL <= 0.016) from Pejaver et al., Am J Hum Genet 2022 (PMID 36413997), placing it in the uninformative interval for PP3/BP4.
• BayesDel noAF score -0.119347 recorded but not used: no verified published threshold is available to this pipeline, so the score is treated as insufficiently calibrated for PP3/BP4 per group instructions.
• aGVGD: no prediction present in the evidence bundle; could not be consulted.

Sources

Reasoning

PP5 applies only to an exact-variant ClinVar expert-panel classification (ClinGen/VCEP expert panel, 3-star review status) of Pathogenic or Likely Pathogenic, per the governing instruction; ordinary laboratory submissions, aggregate ClinVar labels, and non-expert assertions must never trigger it. The ClinVar search for ROS1 c.6068A>T (gene+varname and genomic HGVS queries) returned no variation IDs (variation_ids=[], search_status=no_result, evidence_sentence 'This variant is absent from ClinVar.'). Because no ClinVar record exists for this variant, no expert-panel classification exists to support PP5, and PP5 is therefore not met. This is a definitive negative determination based on a completed ClinVar search, not a data-availability gap.

Rule

pp5_expert_panel_only: requires exact-variant ClinVar expert-panel P/LP classification (3-star); none exists

Evidence considered

• evidence.json clinvar: search_status no_result, variation_ids [], exact_match false (variant absent from ClinVar)
• case_summary.json clinvar.evidence_sentence: 'This variant is absent from ClinVar.'
• prefetch.json clinvar: variation_ids [] and pmids [] across all search terms

Sources

Reasoning

Generic ACMG/AMP 2015 (PMID:25741868) defines BA1 as allele frequency >5% in large population databases (ExAC/gnomAD, 1000 Genomes, ESP). NM_002944.2:c.6068A>T (ROS1, p.Tyr2023Phe) is absent (AF = 0) from gnomAD v2.1 exomes (GRCh37 6-117638373-T-A), gnomAD v4.1 exomes (GRCh38 6-117317210-T-A), and gnomAD-Canada v1.0 genomes. With zero observed alleles in any queried population dataset, the variant does not approach the >5% BA1 threshold, so BA1 is not met. No evidence of a common polymorphism was found in any population source in the case bundle.

Rule

generic_acmg_2015:BA1_af_gt_5_percent

Evidence considered

• gnomAD v2.1 variant page 6-117638373-T-A: absent (search_status 'absent', source_data empty) - AF=0
• gnomAD v4.1 variant page 6-117317210-T-A: absent (search_status 'absent', source_data empty) - AF=0
• gnomAD-Canada v1.0 (HostSeq genomes) API lookup 6-117317210-T-A: absent - AF=0
• BA1 threshold AF>5% per Richards et al. 2015 ACMG/AMP framework (PMID:25741868)

Sources

Reasoning

Generic ACMG/AMP 2015 (PMID:25741868) defines BS1 as an allele frequency greater than expected for the disorder, i.e., too high to be consistent with disease prevalence and penetrance. The variant is absent (AF = 0) from gnomAD v2.1, gnomAD v4.1, and gnomAD-Canada v1.0, so its frequency cannot exceed any expected-disease-frequency threshold; BS1 is therefore not met. Additionally, germline ROS1 disease is not a formally established phenotype (ROS1 is primarily a somatic proto-oncogene; germline context is only tentative familial cancer literature), so no prevalence/penetrance-derived expected AF could be set even if an excess were suspected.

Rule

generic_acmg_2015:BS1_af_exceeds_expected_for_disorder

Evidence considered

• gnomAD v2.1 variant page 6-117638373-T-A: absent - AF=0
• gnomAD v4.1 variant page 6-117317210-T-A: absent - AF=0
• gnomAD-Canada v1.0 (HostSeq genomes) API lookup 6-117317210-T-A: absent - AF=0
• BS1 definition (AF greater than expected for disorder) per Richards et al. 2015 (PMID:25741868)

Still missing

• No established germline disease prevalence/penetrance for ROS1 is available in the bundle, so an expected-allele-frequency upper bound for BS1 could not be derived independently; this does not change the outcome because observed AF=0.

Sources

Reasoning

Generic ACMG/AMP 2015 (PMID:25741868) defines BS2 as observation of the variant in a healthy adult individual (homozygous for recessive, heterozygous for dominant, or hemizygous for X-linked disorders with full penetrance expected at an early age). The variant was not observed at all in gnomAD v2.1, gnomAD v4.1, or gnomAD-Canada v1.0 (no heterozygous or homozygous carriers, no allele counts returned). With no observations of the variant in any healthy/population cohort, the BS2 observation prerequisite is not satisfied; BS2 is not met. Absence in gnomAD does not substitute for a documented healthy-carrier observation, and no such observation appears in the case evidence.

Rule

generic_acmg_2015:BS2_observed_in_healthy_adult

Evidence considered

• gnomAD v2.1 variant page 6-117638373-T-A: absent (no carriers observed)
• gnomAD v4.1 variant page 6-117317210-T-A: absent (no carriers observed)
• gnomAD-Canada v1.0 (HostSeq genomes) API lookup 6-117317210-T-A: absent (no carriers observed)
• BS2 definition (observed in healthy adult, full penetrance at early age) per Richards et al. 2015 (PMID:25741868)

Still missing

• No clinical or cohort data in the bundle documenting the variant in any healthy adult individual; a single well-documented healthy-carrier report could not be checked because none exists in the case evidence.

Sources

Reasoning

BP1 requires a gene in which primarily truncating variants are known to cause disease (missense not a disease mechanism). This is not established for ROS1. The pvs1_gene_context supports loss-of-function as one germline disease mechanism, but nothing in the bundle shows that only truncating variants cause ROS1-associated disease. OncoKB gene-level context states ROS1 is altered by mutation or chromosomal rearrangement across cancers, indicating non-truncating alterations are disease-relevant (somatic oncogenic activation). ROS1 is therefore not a truncating-only gene, and BP1 is not met.

Rule

ACMG/AMP 2015 BP1: Missense variant in a gene for which primarily truncating variants are known to cause disease (Richards et al. 2015, PMID 25741868; generic_acmg_combination_rules).

Evidence considered

• pvs1_gene_context.json: LoF mechanism supported as a germline mechanism (literature-only; familial lung cancer / hereditary breast cancer) - supports LoF but does not exclude missense as a disease mechanism.
• OncoKB (evidence.json oncokb result): 'ROS1, a receptor tyrosine kinase, is altered by mutation or chromosomal rearrangement in a diverse range of cancers, including lung cancer.' - non-truncating alteration mechanisms are disease-relevant.

Still missing

• If a ROS1-specific CSPEC/VCEP or curated gene-level mechanism review emerges, re-evaluate whether a truncating-only disease mechanism can be established.

Sources

Reasoning

BP4 requires multiple lines of computational evidence suggesting no impact. Only one independent calibrated line is available in the benign direction: SpliceAI max delta 0.01 (PMID 30661751) predicts no splice impact. REVEL 0.52 falls in the ClinGen SVI uninformative interval (above the BP4 supporting threshold of <= 0.016, PMID 36413997) and therefore cannot be counted as a benign-direction line. BayesDel -0.119347 is unusable (no verified published threshold). aGVGD is unavailable. Because the missense in-silico path could not contribute a second calibrated line, the 'multiple lines' requirement of BP4 is not satisfied; the criterion is not met.

Rule

Generic ACMG/AMP 2015 (Richards et al., PMID 25741868; source_registry key 'generic_acmg_combination_rules'): BP4 - multiple lines of computational evidence suggest no impact on the gene or gene product (conservation, evolutionary, splicing impact, etc.). No ROS1 CSPEC/VCEP or gene-specific computational lookup spreadsheet exists (framework_mode=generic_acmg, precedence_rank=3).

Evidence considered

• SpliceAI max delta score 0.01 (DS_AG=0.01, all other delta scores 0.00) predicts no splice impact, below the published 0.20/0.50 splice-altering thresholds (Jaganathan et al., Cell 2019, PMID 30661751); counts as a single benign-direction line.
• REVEL score 0.52 is above the ClinGen SVI BP4 supporting threshold (REVEL <= 0.016) from Pejaver et al., Am J Hum Genet 2022 (PMID 36413997); uninformative and not counted toward BP4.
• BayesDel noAF score -0.119347 not counted: no verified published threshold available to this pipeline, treated as insufficiently calibrated per group instructions.
• aGVGD: no prediction present in the evidence bundle; not counted.

Still missing

• Only one independent calibrated computational line (SpliceAI, max delta 0.01) supports no-impact; REVEL 0.52 is in the SVI uninformative interval (PMID 36413997) and BayesDel has no verified published threshold, so the 'multiple lines' requirement of BP4 could not be satisfied.
• aGVGD prediction not present in the evidence bundle; missense-path multi-tool concordance for BP4 could not be evaluated.

Sources

Reasoning

BP6 applies only to an exact-variant ClinVar expert-panel classification (ClinGen/VCEP expert panel, 3-star review status) of Benign or Likely Benign, per the governing instruction; ordinary laboratory submissions, aggregate ClinVar labels, and non-expert assertions must never trigger it. The variant is absent from ClinVar entirely (variation_ids=[], search_status=no_result), so no expert-panel benign classification exists to support BP6. Absence from ClinVar is a lack-of-reporting finding and must not be interpreted as evidence of benignity. BP6 is therefore not met, as a definitive negative determination based on a completed ClinVar search.

Rule

bp6_expert_panel_only: requires exact-variant ClinVar expert-panel B/LB classification (3-star); none exists

Evidence considered

• evidence.json clinvar: search_status no_result, variation_ids [], exact_match false (variant absent from ClinVar)
• case_summary.json clinvar.evidence_sentence: 'This variant is absent from ClinVar.'
• prefetch.json clinvar: variation_ids [] and pmids [] across all search terms

Sources

Reasoning

PS1 requires a different nucleotide change producing the same amino acid change (p.Tyr2023Phe) that is established as pathogenic. The variant is absent from ClinVar (no record, no submissions) and literature_pass.json contains 0 PMIDs, so no same-amino-acid comparator could be identified. The pm5_candidates harness ran in 'unknown' mode and explicitly skipped automatic candidate harvesting (total candidates seen = 0), so an alternate-change search at codon 2023 was not actually performed. PS1 cannot be confirmed or excluded on the available evidence.

Rule

ACMG/AMP 2015 PS1: Same amino acid change as a previously established pathogenic variant regardless of nucleotide change (Richards et al. 2015, PMID 25741868; generic_acmg_combination_rules).

Evidence considered

• ClinVar search (ROS1[GENE] AND c.6068A>T[VARNAME] and related HGVS terms): no record found (evidence.json clinvar result, search_status no_result).
• pm5_candidates artifact: mode 'unknown', recommendation level 'not_applicable', total candidates seen 0, same-residue candidates 0 (case_summary.json pm5_candidates_summary).
• literature_pass.json: total_pmids 0, criteria_index empty.

Still missing

• No search was performed for alternate nucleotide changes at codon 2023 that produce p.Tyr2023Phe (e.g., a TAC->TTC change at c.6069); ClinVar has no entry for c.6068A>T and the literature pass returned 0 PMIDs, so an established-pathogenic same-amino-acid comparator cannot be confirmed or excluded.

Sources

Reasoning

PS2 (de novo, paternity and maternity confirmed) cannot be evaluated: the case-evidence bundle contains no proband phenotype, no family history, and no parental genotyping/trio data for NM_002944.2:c.6068A>T. ClinVar has no record for this variant (evidence.json: 'This variant is absent from ClinVar.'), the literature pass returned 0 PMIDs, and the publications/ directory is empty, so no confirmed de novo occurrence has been reported in any available source. With no trio/parental-testing evidence whatsoever, the criterion is not_assessed rather than not_met.

Rule

ACMG/AMP 2015 (PMID 25741868), PS2: De novo (paternity and maternity confirmed) variant in a patient with the disorder and no family history. Generic fallback (no ROS1 VCEP/CSPEC).

Evidence considered

• Richards S et al. 2015, Standards and guidelines for the interpretation of sequence variants, Genet Med 17(5):405-424 (PMID 25741868) — generic ACMG/AMP PS2 definition: de novo variant with confirmed paternity and maternity; applied because no ROS1 CSPEC/VCEP specification exists.
• ClinVar search for NM_002944.2:c.6068A>T (ROS1[GENE] AND c.6068A>T[VARNAME] and alternate query terms) returned no record, confirming no submitted de novo observation.

Still missing

• No proband de novo testing data: the bundle contains no proband genotype result, parental genotype results, or maternity/paternity confirmation for NM_002944.2:c.6068A>T.
• No clinical phenotype/family-history context for the proband is present in the case bundle, so phenotype consistency with a ROS1-associated germline disorder cannot be evaluated.
• No published or ClinVar-submitted de novo observation of c.6068A>T exists (0 PMIDs; ClinVar no_result).

Sources

Reasoning

Generic ACMG/AMP 2015 framework applies (no ROS1 CSPEC/VCEP specification exists; cspec not found, framework_mode=generic_acmg, precedence_rank=3). PS3 requires well-established in vitro or in vivo functional studies demonstrating a damaging effect on the gene or gene product. No such evidence exists for NM_002944.2:c.6068A>T (p.Tyr2023Phe): the literature pass retrieved zero PMIDs, the publications directory is empty, and fulltext_map is empty. OncoKB reports 'Unknown Oncogenic Effect' and explicitly notes no variant-specific reviewed functional evidence for this variant. SpliceAI (max delta score 0.01) predicts no splice impact, but this is an in silico prediction, not a functional assay, and cannot substitute for PS3 evidence. Because no functional data are available at all, PS3 can be neither met nor formally excluded; per instruction, insufficient evidence is recorded as not_assessed.

Rule

PS3 - Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product (Richards et al. 2015, PMID 25741868; generic ACMG/AMP fallback, no ROS1 VCEP/CSPEC available)

Evidence considered

• Generic ACMG/AMP 2015 rules (generic_acmg_combination_rules; source_pmid 25741868): PS3 applies when well-established functional studies support a damaging effect on the gene or gene product.
• OncoKB (oncokb): classification 'Unknown Oncogenic Effect'; is_variant_specific=false, not_reviewed=true — no variant-specific functional evidence exists to support a damaging effect.
• SpliceAI (spliceai): max delta score 0.01, all DS scores <=0.01 — no predicted splice impact; in silico prediction only, not a well-established functional study, and therefore not PS3 evidence.

Still missing

• No in vitro or in vivo functional assay data for ROS1 p.(Tyr2023Phe) (c.6068A>T) identified in any reviewed source.
• Literature pass retrieved zero PMIDs; no variant-specific functional publications were available to review.
• OncoKB classifies the variant as 'Unknown Oncogenic Effect' with no variant-specific functional evidence (not_reviewed=true), so no curated functional conclusion can be adopted.

Sources

Reasoning

PS4 (Richards et al. 2015, PMID:25741868) requires statistically significant enrichment of the variant in affected individuals versus matched unaffected controls in case-control or cohort studies. No such data exist in this case evidence bundle: literature_pass.json reports total_pmids=0 with no high- or medium-priority papers, fulltext_map.json is empty, and no cohort/case-control dataset appears in case_summary.json or evidence.json. The variant is absent from gnomAD v2.1, gnomAD v4.1, and gnomAD-Canada v1.0, documenting rarity in population databases, but population absence is not a case-control frequency comparison and cannot support PS4 on its own. Without any affected-vs-unaffected allele-frequency comparison for c.6068A>T, PS4 cannot be evaluated.

Rule

acmg_ps4: case-control enrichment in affected vs unaffected cohorts (PMID:25741868)

Evidence considered

• literature_pass.json: total_pmids=0, no high/medium priority papers, criteria_index empty
• evidence.json gnomad: search_status absent for gnomAD v2.1 (6-117638373-T-A), v4.1 (6-117317210-T-A), and gnomAD-Canada v1.0
• generic_acmg_classification_rules.md: generic ACMG/AMP 2015 fallback (PMID:25741868) is the governing framework (cspec found=false)

Still missing

• No case-control or cohort allele-frequency comparison of c.6068A>T in affected individuals versus matched unaffected controls
• No published case series or cohort study reporting this exact variant
• Variant-specific literature search returned no PMIDs (literature_pass total_pmids=0)

Sources

Reasoning

PM5 requires a different missense change at the same residue (p.Tyr2023) that has been established as pathogenic. The dedicated pm5_candidates artifact ran in 'unknown' mode, recommended 'not_applicable', and skipped classic same-residue candidate harvesting; it collected 0 candidates (0 same-residue, 0 expert-panel PLP, 0 conflicted). ClinVar has no entry for this variant and literature_pass.json contains no PMIDs. Following the artifact's hermes_instruction to review whether any collected same-residue comparators satisfy the PM5 rule: none were collected, so PM5 cannot be applied. Because the harvesting search itself was skipped rather than completed, absence of a comparator is unverified and PM5 is left not_assessed rather than met.

Rule

ACMG/AMP 2015 PM5: Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before (Richards et al. 2015, PMID 25741868; generic_acmg_combination_rules).

Evidence considered

• pm5_candidates artifact: recommendation level 'not_applicable', summary 'Unable to confirm classic same-residue PM5 semantics safely; skipping automatic candidate harvesting.'; candidate_summary total_candidates_seen 0 (case_summary.json pm5_candidates_summary).
• ClinVar search for c.6068A>T: no record found (evidence.json clinvar result).
• literature_pass.json: total_pmids 0 - no publications screenable for same-residue variants.

Still missing

• pm5_candidates harness mode 'unknown' - classic same-residue PM5 harvesting was skipped; zero candidates collected. No ClinVar entry and no literature PMIDs, so a pathogenic different-missense comparator at p.Tyr2023 cannot be confirmed or excluded without a dedicated same-residue literature/ClinVar search.

Sources

Reasoning

PM6 (assumed de novo without confirmation of paternity/maternity) cannot be evaluated: no de novo occurrence of c.6068A>T is reported anywhere in the case bundle, and no parental testing results are available to support or refute an assumed de novo origin. The variant is absent from ClinVar and from all literature retrieved (0 PMIDs). Without any proband/family data, PM6 is not_assessed.

Rule

ACMG/AMP 2015 (PMID 25741868), PM6: Assumed de novo, but without confirmation of paternity and maternity. Generic fallback (no ROS1 VCEP/CSPEC).

Evidence considered

• Richards S et al. 2015, Genet Med 17(5):405-424 (PMID 25741868) — generic ACMG/AMP PM6 definition: assumed de novo, without confirmation of paternity and maternity.
• ClinVar search for NM_002944.2:c.6068A>T returned no record; literature_pass.json total_pmids = 0, so no assumed-de novo case report is available.

Still missing

• No parental testing data (affected proband plus unaffected-parent genotyping) is available, which is the minimum required to assess an assumed de novo origin.
• No report of the variant occurring de novo in any individual exists in ClinVar or the retrieved literature (0 PMIDs).
• No proband clinical information is provided in the case bundle, precluding assessment of phenotype consistency for a de novo event.

Sources

Reasoning

PP1 (co-segregation with disease in multiple affected family members) cannot be evaluated: the case bundle contains no pedigree, no affected-family-member genotypes, and no segregation data for c.6068A>T. The variant has no ClinVar record and no literature reports (0 PMIDs), so no segregation study of this variant exists in any available source. Additionally, no established germline disease phenotype for ROS1 missense variants is provided in the bundle, so affected-status consistency cannot be assessed. PP1 is therefore not_assessed.

Rule

ACMG/AMP 2015 (PMID 25741868), PP1: Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease. Generic fallback (no ROS1 VCEP/CSPEC).

Evidence considered

• Richards S et al. 2015, Genet Med 17(5):405-424 (PMID 25741868) — generic ACMG/AMP PP1 definition: co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
• ClinVar search for NM_002944.2:c.6068A>T returned no record; literature_pass.json total_pmids = 0, confirming absence of any segregation report for this variant.

Still missing

• No segregation data: no affected family members have been genotyped for c.6068A>T, and no pedigree is provided in the case bundle.
• No published or ClinVar-reported segregation of this variant (0 PMIDs; ClinVar no_result).
• No ROS1-associated germline disease phenotype is defined in the bundle, so cosegregation with a consistent phenotype cannot be evaluated.

Sources

Reasoning

PP2 requires (a) a low rate of benign missense variation in the gene and (b) missense as a common mechanism of disease. No ROS1 missense-constraint data (gnomAD missense Z-score / missense o/e) are present anywhere in the case bundle, and no source in the bundle establishes that missense variants are a common mechanism of ROS1-associated germline disease. The only gene-level disease-mechanism context available (pvs1_gene_context) supports loss-of-function as a germline mechanism (familial lung cancer / hereditary breast cancer literature), which addresses LoF, not missense. Both PP2 requirements are therefore unassessable from the available evidence.

Rule

ACMG/AMP 2015 PP2: 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 (Richards et al. 2015, PMID 25741868; generic_acmg_combination_rules).

Evidence considered

• pvs1_gene_context.json: germline disease context found via literature only (familial lung cancer risk GELCC, hereditary breast cancer); lof_mechanism_supported true - addresses LoF mechanism only, no missense-mechanism statement.
• OncoKB gene-level context: ROS1 altered by mutation or chromosomal rearrangement in cancers (somatic context); no germline missense disease-mechanism statement.

Still missing

• No gnomAD missense constraint (missense Z-score or missense o/e) data for ROS1 in the bundle; no evidence establishing missense as a common mechanism of ROS1-associated germline disease. Obtain constraint data from a named source and gene-level mechanism evidence before PP2 can be scored.

Sources

Reasoning

PP4 (Richards et al. 2015, PMID:25741868) supports pathogenicity when the phenotype of a variant carrier is highly specific for the gene's disease, typically in the absence of case-control data. No proband phenotype, clinical case description, or case series is available for this variant: literature_pass.json contains 0 PMIDs, and no phenotype information appears in case_summary.json or evidence.json. The only variant-specific curated context is OncoKB's classification of 'Unknown Oncogenic Effect' with is_variant_specific=false and no reviewed clinical phenotype evidence for ROS1 p.Tyr2023Phe. Without any documented phenotype associated with a carrier of c.6068A>T, phenotype specificity cannot be evaluated.

Rule

acmg_pp4: phenotype highly specific for gene disease in absence of case-control data (PMID:25741868)

Evidence considered

• evidence.json oncokb: classification 'Unknown Oncogenic Effect', is_variant_specific=false, no variant-specific clinical phenotype evidence
• literature_pass.json: total_pmids=0 (no phenotype or case literature available)
• case_summary.json: no proband/phenotype fields present in the evidence bundle

Still missing

• No clinical phenotype data for any carrier of c.6068A>T
• No case report or case series describing a phenotype associated with this exact variant
• No ROS1 VCEP/CSPEC phenotype-specific criteria specification (cspec found=false; generic ACMG fallback)

Sources

Reasoning

Generic ACMG/AMP 2015 framework applies (no ROS1 CSPEC/VCEP specification exists). BS3 requires well-established functional studies showing no damaging effect on protein function or splicing. No such studies exist for NM_002944.2:c.6068A>T (p.Tyr2023Phe): zero PMIDs from the literature pass, empty publications directory, variant absent from ClinVar, and OncoKB reports 'Unknown Oncogenic Effect' with no variant-specific functional evidence. The low SpliceAI max delta score (0.01) is an in silico prediction and is not a well-established functional study; per ACMG/AMP it cannot support BS3. Absence of functional evidence cannot be equated with evidence of a benign effect, so BS3 is not met; because no functional data are available at all, the criterion is recorded as not_assessed rather than not_met.

Rule

BS3 - Well-established functional studies show no damaging effect on protein function or splicing (Richards et al. 2015, PMID 25741868; generic ACMG/AMP fallback, no ROS1 VCEP/CSPEC available)

Evidence considered

• Generic ACMG/AMP 2015 rules (generic_acmg_combination_rules; source_pmid 25741868): BS3 applies when well-established functional studies show no damaging effect on protein function or splicing.
• OncoKB (oncokb): 'Unknown Oncogenic Effect'; is_variant_specific=false, not_reviewed=true — provides no functional evidence of a benign effect.
• SpliceAI (spliceai): max delta score 0.01 — in silico only, insufficient to establish no damaging effect under BS3.
• ClinVar (clinvar): variant absent from ClinVar (no submissions), so no curated functional or benign classification evidence is available.

Still missing

• No functional assay demonstrating normal function or absence of a damaging effect for p.(Tyr2023Phe) (c.6068A>T).
• In silico splice prediction (SpliceAI max delta 0.01) is not a well-established functional study and cannot satisfy BS3.
• No ClinVar submissions, literature reports, or curated resources (OncoKB 'Unknown Oncogenic Effect') provide any functional evidence of a benign effect.

Sources

Reasoning

BS4 (lack of segregation in affected family members) cannot be evaluated: no non-segregation data exist for this variant. The bundle contains no family genotypes or affected-relative testing results, and the variant is unreported in ClinVar and in all retrieved literature (0 PMIDs), so no study has tested segregation of c.6068A>T. Absence of a variant from population databases (gnomAD) is a frequency observation and does not substitute for BS4 segregation evidence under generic ACMG/AMP. BS4 is therefore not_assessed.

Rule

ACMG/AMP 2015 (PMID 25741868), BS4: Lack of segregation in affected family members. Generic fallback (no ROS1 VCEP/CSPEC).

Evidence considered

• Richards S et al. 2015, Genet Med 17(5):405-424 (PMID 25741868) — generic ACMG/AMP BS4 definition: lack of segregation in affected family members.
• ClinVar search for NM_002944.2:c.6068A>T returned no record; literature_pass.json total_pmids = 0, confirming no segregation or non-segregation observations for this variant.

Still missing

• No non-segregation evidence: no affected relatives tested and found not to carry c.6068A>T, and no family segregation studies are available.
• No pedigree or family testing data is present in the case bundle.
• No literature or ClinVar report of this variant exists against which segregation could be checked (0 PMIDs; ClinVar no_result).

Sources

Reasoning

BP2 (Richards et al. 2015, PMID 25741868) is a benign criterion requiring an actual allele-phase observation: the variant observed in trans with a pathogenic variant for a fully dominant disease/mechanism, or observed in cis with a pathogenic variant in any inheritance pattern. No such observation exists for ROS1 c.6068A>T. The variant is absent from ClinVar (no submissions), absent from gnomAD v2.1/v4.1/Canada, absent from COSMIC, and no variant-specific publications were identified (literature_pass total = 0). No proband, pedigree, segregation, or phase (trans/cis) data are available in the case bundle, and no pathogenic variant has been reported in trans or cis with this variant in any source. The criterion cannot be evaluated from the available evidence, so it is recorded as not_assessed rather than met or not_met; it contributes no benign scoring evidence. A human reviewer should confirm whether any external family or phase data exist outside this bundle.

Rule

BP2 (Richards et al. 2015, PMID 25741868, generic ACMG/AMP fallback via generic_acmg_combination_rules): 'Observed in trans with a pathogenic variant for a fully dominant disease/mechanism; or observed in cis with a pathogenic variant in any inheritance pattern.'

Evidence considered

• BP2 rule text (Richards et al. 2015, PMID 25741868): 'Observed in trans with a pathogenic variant for a fully dominant disease/mechanism; or observed in cis with a pathogenic variant in any inheritance pattern' — requires a documented phase observation that is entirely absent here.
• clinvar: 'This variant is absent from ClinVar.' — no submissions reporting the variant, so no trans/cis phase observations can be derived.
• pvs1_gene_context: germline context is 'literature_only' with empty inheritance list — no established disease/mechanism against which a phase observation could be interpreted.

Still missing

• No proband, pedigree, segregation, or allele-phase (trans/cis) data are available for this variant in any source consulted.
• No pathogenic variant has been reported in trans or cis with c.6068A>T; no ClinVar submissions or variant-specific literature exist to provide such observations.
• ROS1 germline disease context is literature-only with no established fully dominant mechanism, further limiting the applicability assessment of the BP2 dominant branch.

Sources

Reasoning

BP5 (Richards et al. 2015, PMID:25741868) applies when a variant is found in a case with an alternative molecular basis for disease, arguing that the variant is not the cause. Evaluation requires proband-level genetic/clinical data demonstrating a distinct causative alteration. No proband, family, or case-level data are present in this bundle: the variant has no ClinVar record, literature_pass.json reports 0 PMIDs, and no case-level genetic findings appear in case_summary.json or evidence.json. Without any proband-level information, the existence or absence of an alternative molecular basis cannot be evaluated.

Rule

acmg_bp5: variant found in case with an alternative molecular basis for disease (PMID:25741868)

Evidence considered

• literature_pass.json: total_pmids=0; no case-level genetic data in the evidence bundle
• evidence.json clinvar: no ClinVar record (no proband submissions or co-occurrence data)
• generic_acmg_classification_rules.md: generic ACMG/AMP 2015 fallback (PMID:25741868) is the governing framework

Still missing

• No proband-level clinical or genetic data to determine whether an alternative molecular basis for disease exists
• No family segregation or co-occurrence data for c.6068A>T
• Variant-specific literature search returned no PMIDs (literature_pass total_pmids=0)

Sources