# A Comparison of Variant Interpretation Guidelines: ACMG/AMP vs. ClinGen vs. AMP/ASCO/CAP for Somatic Variants ## Key Takeaways - The ACMG/AMP framework is designed for germline variants in Mendelian diseases, focusing on hereditary risk and utilizing 28 criteria across pathogenic and benign categories, while the AMP/ASCO/CAP tier system is for somatic variants in cancer, prioritizing clinical actionability into four tiers (I-IV) based on therapeutic, diagnostic, or prognostic relevance. - ClinGen provides gene-specific specifications that refine and calibrate the ACMG/AMP criteria, often incorporating quantitative Bayesian-informed scoring and novel evidence types like variant allele fraction in specific contexts (e.g., clonal hematopoiesis), aiming to reduce VUS rates and improve inter-laboratory concordance. - Somatic variant interpretation relies on evidence of therapeutic actionability from sources like FDA drug labels and clinical trials, alongside diagnostic/prognostic data, and critically requires distinguishing somatic from germline alterations through paired tumor-normal analysis or explicit filtering based on variant allele frequency and population databases. - Applying germline criteria (ACMG/AMP) to somatic variants leads to misleading classifications due to differing evidence standards, while applying somatic criteria (AMP/ASCO/CAP) to germline variants misses crucial hereditary risk information; therefore, the testing context (sample type and clinical question) dictates the appropriate framework. - Common failure patterns include misapplying germline criteria to somatic variants, confusing somatic with germline alterations (especially in liquid biopsies), ignoring clonal hematopoiesis of indeterminate potential (CHIP) as a source of false somatic calls, and overinterpreting variants of uncertain significance (VUS) or Tier III variants without sufficient evidence. --- Clinical molecular pathologists and oncology laboratory professionals face a practical decision when interpreting genomic variants: which classification framework applies to their testing context. Germline variant interpretation follows the American College of Medical Genetics and Genomics and Association for Molecular Pathology (ACMG/AMP) framework, while somatic variant interpretation in cancer follows the AMP/ASCO/CAP tier system. ClinGen provides gene-specific specifications and infrastructure that refine both approaches. This article compares these frameworks, explains their evidence structures, and provides concrete criteria for selecting the appropriate system based on sample type, clinical question, and reporting requirements. ## Scope and Reader Context The distinction between germline and somatic variant interpretation reflects fundamental differences in biological meaning, clinical actionability, and evidence standards. Germline variants are inherited or arise de novo in reproductive cells, are present in essentially all tissues, and inform hereditary cancer risk, carrier status, and pharmacogenetic decisions. Somatic variants arise during an individual's lifetime in specific tissues, are present only in the tumor or affected tissue, and inform tumor biology, targeted therapy selection, prognosis, and treatment resistance. Laboratory professionals must determine which framework applies before designing a variant calling workflow, selecting filtering strategies, or assigning clinical significance. Applying germline criteria to somatic variants produces misleading classifications because the evidence types differ. Applying somatic criteria to germline variants misses hereditary risk information that may have implications for family members. The practical question for most laboratories is which framework matches the clinical question and sample type. This article provides a side-by-side comparison of the ACMG/AMP framework, ClinGen specifications, and the AMP/ASCO/CAP tier system, with emphasis on evidence tiers, actionability, therapeutic implications, and workflow decisions. ## At a Glance: Framework Comparison | Feature | ACMG/AMP | ClinGen | AMP/ASCO/CAP | |---------|----------|---------|--------------| | Primary context | Germline variants in Mendelian disease genes | Gene-specific refinement of ACMG/AMP criteria | Somatic variants in cancer specimens | | Evidence categories | 28 criteria across pathogenic (PVS1, PS1-PS4, PM1-PM6, PP1-PP5) and benign (BA1, BS1-BS4, BP1-BP7) | Modified criteria with gene-specific strength adjustments and quantitative likelihood ratio approaches | Four tiers: Tier I (strong clinical significance), Tier II (potential clinical significance), Tier III (unknown significance), Tier IV (benign or likely benign) | | Key evidence types | Population frequency, segregation, functional studies, de novo status, allelic data, computational prediction | Variant allele fraction, clonal hematopoiesis context, gene-specific population thresholds, quantitative Bayesian-informed scoring | Variant allele frequency, paired tumor-normal comparison, therapeutic actionability, prognostic databases, pathway involvement | | Actionability focus | Hereditary risk, reproductive decision-making, surveillance | Hereditary risk with gene-specific nuance | Therapeutic targeting, resistance mechanisms, diagnostic classification | | Reporting output | Pathogenic, likely pathogenic, VUS, likely benign, benign | Same five-tier system with gene-specific calibration | Tier I-IV with therapeutic, diagnostic, or prognostic relevance | | Primary databases | ClinVar, gnomAD, locus-specific databases | ClinGen Variant Curation Expert Panels, CSpec registry | COSMIC, OncoKB, CIViC, ClinVar | | Best use case | Constitutional testing, carrier screening, hereditary cancer panels | Germline variants in genes with expert panel specifications | Tumor profiling, liquid biopsy, treatment selection | ## Core Principles of Variant Classification ### Biological Distinction Between Germline and Somatic Variants Germline variants are present in every nucleated cell of an individual. They are inherited from parents or arise as de novo mutations in gametes or early embryonic development. The clinical question for germline variants is whether the variant increases risk for disease, and if so, what surveillance or management actions are appropriate. The ACMG/AMP framework was designed for this context, with evidence categories that weigh population frequency, segregation in affected families, functional impact, and established disease mechanism. Somatic variants are acquired mutations in somatic cells. They are present only in the tumor or affected tissue and are not transmitted to offspring. The clinical question for somatic variants is whether the mutation drives tumor growth, whether it predicts response to specific therapies, and whether it provides diagnostic or prognostic information. The AMP/ASCO/CAP tier system was designed for this context, with tiers based on clinical actionability instead of hereditary pathogenicity. A practical example illustrates the distinction. A TP53 variant detected in a tumor sample could be somatic, germline, or associated with clonal hematopoiesis of indeterminate potential (CHIP). The interpretation depends on whether the variant is present in matched normal tissue, the variant allele fraction, and the gene involved. The ClinGen TP53 Variant Curation Expert Panel has updated its specifications to incorporate variant allele fraction as evidence of pathogenicity, particularly in the context of clonal hematopoiesis. This refinement acknowledges that the same gene and variant may require different interpretive frameworks depending on the tissue context. ### Evidence Standards and Their Limitations The ACMG/AMP framework uses a combinatorial scoring system. Each variant is evaluated against 28 criteria, and the combination of pathogenic and benign criteria determines the final classification. The framework was designed to standardize interpretation across laboratories, but it has known limitations. Variants of uncertain significance (VUS) remain common, particularly in genes with limited population data or unclear disease mechanism. The ClinGen TP53 Expert Panel reported that many TP53 variants remain classified as VUS, and their updated specifications were designed to reduce VUS rates and increase inter-laboratory concordance. The AMP/ASCO/CAP tier system uses a different evidence structure. Instead of combinatorial criteria, it assigns tiers based on the clinical actionability of the variant. Tier I variants have strong clinical significance and include FDA-approved therapy targets, resistance mechanisms, and diagnostic or prognostic markers with professional guideline support. Tier II variants have potential clinical significance and include investigational therapy targets or variants with limited but suggestive evidence. Tier III variants have unknown clinical significance, and Tier IV variants are benign or likely benign. The evidence standards differ in their treatment of functional data. The ACMG/AMP framework includes functional studies as evidence (PS3 for pathogenic, BS3 for benign), but the strength of this evidence depends on whether the functional assay accurately reflects the disease mechanism. The AMP/ASCO/CAP framework places greater emphasis on therapeutic actionability, with evidence drawn from clinical trials, drug labels, and professional guidelines. ## The ACMG/AMP Framework for Germline Variants ### Evidence Categories and Scoring The ACMG/AMP framework assigns variants to one of five classes: pathogenic, likely pathogenic, variant of uncertain significance, likely benign, or benign. The classification uses a defined set of evidence criteria, each with specified strength levels. Pathogenic criteria include PVS1 (null variant in a gene where loss of function is the disease mechanism), PS1-PS4 (same amino acid change as an established pathogenic variant, de novo status, functional studies, and segregation), PM1-PM6 (mutational hotspot, absent in controls, recessive disease detection, protein length change, novel missense in a gene with low benign missense rate, and assumed de novo without confirmation), and PP1-PP5 (co-segregation, patient phenotype, family history, and reputable source). Benign criteria include BA1 (allele frequency too high for the disorder), BS1-BS4 (allele frequency greater than expected, observed in healthy adults, lack of segregation, and lack of functional effect), and BP1-BP7 (missense in a gene where truncating variants cause disease, observed in trans with a pathogenic variant, in-frame indels in repetitive regions, multiple observations with no disease, missense in a gene with high benign missense rate, synonymous or intronic variant, and computational prediction without functional evidence). The framework requires careful application of each criterion. For example, PVS1 applies only when loss of function is the established disease mechanism for the gene. Applying PVS1 to a gene where gain of function or dominant-negative mechanisms cause disease would produce an incorrect classification. Similarly, BA1 requires a population frequency threshold that exceeds the disease prevalence, and the threshold varies by gene and disorder. ### ClinGen Gene-Specific Specifications ClinGen addresses the limitations of the generic ACMG/AMP framework through Variant Curation Expert Panels (VCEPs). These panels develop gene-specific specifications that modify the strength of evidence criteria based on gene-specific data. The ClinGen TP53 Variant Curation Expert Panel provides a concrete example. Their updated specifications incorporate the latest ClinGen recommendations and methodological advances, providing greater granularity for multiple evidence types. They introduced the novel use of variant allele fraction as evidence of pathogenicity, particularly in the context of clonal hematopoiesis. The TP53 panel followed a data-driven approach using likelihood ratio-based quantitative analyses to guide code application and determine strength modifications. This Bayesian-informed approach allows the panel to calibrate evidence strength based on quantitative data instead of expert judgment alone. The panel reported that the updated specifications decreased VUS rates and increased certainty, with clinically meaningful classifications for 93% of pilot variants. ClinGen also provides infrastructure for variant classification through its application programming interface-based microservices. These services expose findable, accessible, interoperable, reusable, and AI-ready variant data, laying a foundation for next-generation software applications, AI systems, and variant classification workflows. For laboratory professionals, this means that ClinGen specifications and variant data can be integrated into local variant interpretation pipelines. ### Practical Application for Germline Testing Laboratories performing germline testing should determine whether a ClinGen VCEP specification exists for each gene on their panel. The ClinGen Criteria Specifications Registry (CSpec) provides the most current version of each specification. When a specification exists, the laboratory should apply the gene-specific criteria instead of the generic ACMG/AMP framework. When no specification exists, the generic framework applies, with careful attention to gene-specific disease mechanisms. The TP53 example illustrates the practical impact of gene-specific specifications. The updated TP53 specifications are expected to reduce VUS rates, increase inter-laboratory concordance, and improve medical management for individuals with germline TP53 variants. For a laboratory reporting TP53 germline variants, using the outdated generic criteria would produce more VUS classifications and potentially miss clinically meaningful classifications. ## The AMP/ASCO/CAP Tier System for Somatic Variants ### Tier Structure and Clinical Actionability The AMP/ASCO/CAP tier system classifies somatic variants based on their clinical significance in cancer. The system uses four tiers, each defined by the strength of clinical actionability evidence. Tier I variants have strong clinical significance. These include variants in genes with FDA-approved therapies, variants that predict resistance to approved therapies, and variants with diagnostic or prognostic significance recognized by professional guidelines. Examples include EGFR activating mutations in non-small cell lung cancer, which predict response to EGFR tyrosine kinase inhibitors, and KRAS mutations that predict resistance to EGFR inhibitors in colorectal cancer. Tier II variants have potential clinical significance. These include variants in genes that are targets of investigational therapies in clinical trials, variants with limited but suggestive evidence of therapeutic relevance, and variants with diagnostic or prognostic significance based on smaller studies or case reports. The distinction between Tier I and Tier II depends on the maturity of the evidence, not the biological importance of the variant. Tier III variants have unknown clinical significance. These include variants in cancer-related genes without sufficient evidence to assign therapeutic, diagnostic, or prognostic relevance. Tier III variants are analogous to VUS in the germline framework but are specific to the somatic context. Tier IV variants are benign or likely benign. These include common polymorphisms and variants with no known role in cancer. ### Evidence Sources for Somatic Classification The AMP/ASCO/CAP framework relies on evidence from multiple sources. Therapeutic actionability evidence comes from FDA drug labels, National Comprehensive Cancer Network guidelines, clinical trial databases, and curated knowledge bases. Diagnostic and prognostic evidence comes from professional guidelines, large cohort studies, and disease-specific literature. The framework also considers variant allele frequency and the presence of the variant in matched normal tissue. A variant present at high allele frequency in a tumor sample but absent in matched normal tissue is likely somatic. A variant present at similar allele frequency in both tumor and normal tissue may be germline. A variant present at low allele frequency in normal tissue may represent clonal hematopoiesis. The Cancer of Unknown Primary (CUP) study provides a practical example of somatic variant classification in liquid biopsy. The study applied a 92-gene CUP-specific targeted sequencing panel to liquid biopsy samples from 39 CUP patients, analyzing circulating cell-free DNA together with paired germline DNA when available. Somatic, germline, and CHIP-related variants were classified using predefined variant allele frequency thresholds and paired ccfDNA/gDNA comparison. The study identified 44 clinically relevant variants (Tier I-III), most frequently affecting NF1, KRAS, ARID1A, and PIK3CA. Actionable mutations were identified in 13 genes, supporting the potential role of molecularly guided therapies in CUP. This example illustrates several practical points. First, paired germline DNA analysis is essential for distinguishing somatic from germline variants. Second, variant allele frequency thresholds help identify CHIP-related variants. Third, the tier system provides a framework for reporting clinically relevant variants even when the primary tumor site is unknown. ### Somatic Variant Calling Workflow The somatic variant calling workflow differs from germline variant calling in several respects. Somatic variant calling requires matched normal tissue to distinguish somatic from germline variants. When matched normal tissue is unavailable, the laboratory must rely on variant allele frequency, population databases, and gene-specific knowledge to make this distinction. The workflow begins with sequencing data from tumor and normal samples. Alignment and preprocessing follow standard bioinformatics practices. Somatic variant callers identify variants present in the tumor but absent or at lower frequency in the normal sample. The variant caller output is filtered based on quality metrics, variant allele frequency, and genomic context. Variant filtering for somatic analysis differs from germline analysis. Somatic variant callers typically use lower allele frequency thresholds because tumor samples may contain heterogeneous cell populations. The filtering strategy must account for tumor purity, copy number alterations, and sequencing artifacts. The CUP study used predefined variant allele frequency thresholds and paired ccfDNA/gDNA comparison to classify variants, demonstrating the importance of explicit filtering criteria. After variant calling and filtering, the laboratory assigns each variant to a tier based on clinical actionability. This step requires access to curated knowledge bases and professional guidelines. The laboratory must document the evidence supporting each tier assignment and the version of the knowledge base used. ## ClinGen Infrastructure and Its Role in Both Frameworks ### Variant Curation Expert Panels ClinGen VCEPs develop gene-specific specifications for both germline and somatic variant interpretation. While the primary focus has been germline variants, the infrastructure and methods apply to somatic variants as well. The TP53 VCEP provides an example of how gene-specific specifications can incorporate somatic context, including variant allele fraction and clonal hematopoiesis. The VCEP process involves multiple steps. The panel reviews the gene-specific disease mechanism, population frequency data, functional assay evidence, and clinical data. The panel then modifies the ACMG/AMP criteria to reflect gene-specific evidence strengths. Proposed modifications are discussed in working group meetings and subjected to comprehensive review during monthly general VCEP meetings to reach consensus. The TP53 VCEP used likelihood ratio-based quantitative analyses to guide code application and determine strength modifications. This approach allows the panel to assign quantitative weights to evidence types based on their predictive value. The panel reported that the updated specifications led to both decreased VUS and increased certainty, with clinically meaningful classifications for 93% of variants. ### CSpec Registry and API Infrastructure The ClinGen Criteria Specifications Registry (CSpec) provides the most current version of each VCEP specification. Laboratory professionals should check the CSpec registry before applying gene-specific criteria. The registry provides versioned specifications, allowing laboratories to track changes over time. ClinGen's application programming interface-based microservices accelerate growth and dissemination of knowledge about human genetic variation. By exposing findable, accessible, interoperable, reusable, and AI-ready variant data, ClinGen lays a foundation for next-generation software applications, AI systems, and variant classification workflows. For laboratory professionals, this means that ClinGen specifications and variant data can be integrated into local variant interpretation pipelines. ### Integration With Laboratory Workflows Laboratories can integrate ClinGen specifications into their variant interpretation workflows in several ways. The CSpec registry provides the authoritative version of each specification. The API infrastructure allows programmatic access to variant data and specifications. The VCEP publications provide the rationale and evidence supporting each specification. The practical integration depends on the laboratory's variant interpretation platform. Some platforms include ClinGen specifications as part of their knowledge base. Others require manual application of the specifications. The laboratory should document which specification version was used for each variant classification and when the specification was updated. ## Practical Implementation Steps ### Step 1: Determine the Testing Context Before selecting a variant interpretation framework, the laboratory must determine the testing context. Germline testing for hereditary cancer risk requires the ACMG/AMP framework with ClinGen specifications where available. Somatic testing for tumor profiling requires the AMP/ASCO/CAP tier system. Some tests, such as tumor-normal paired testing, may require both frameworks. The testing context determines the sample requirements, the variant calling workflow, and the reporting format. Germline testing typically uses blood, saliva, or buccal samples. Somatic testing uses tumor tissue or liquid biopsy samples. Paired tumor-normal testing requires both sample types. ### Step 2: Design the Variant Calling Workflow The variant calling workflow must match the testing context. Germline variant calling uses germline variant callers and typically requires higher allele frequency thresholds. Somatic variant calling uses somatic variant callers and requires matched normal tissue or explicit filtering criteria for unmatched samples. The workflow should include quality control steps at each stage. Sequence alignment quality, coverage depth, and variant caller concordance should be monitored. The laboratory should document the pipeline version and parameters used for each analysis. ### Step 3: Apply the Appropriate Classification Framework For germline variants, apply the ACMG/AMP framework with ClinGen specifications where available. Check the CSpec registry for the most current specification version. Apply gene-specific criteria instead of generic criteria when a specification exists. For somatic variants, apply the AMP/ASCO/CAP tier system. Assign each variant to a tier based on clinical actionability evidence. Document the evidence supporting each tier assignment and the knowledge base version used. ### Step 4: Report Results With Appropriate Context The report should clearly indicate whether the variant interpretation used the germline or somatic framework. For somatic variants, the report should include the tier assignment and the clinical actionability evidence. For germline variants, the report should include the five-tier classification and the evidence supporting the classification. The report should also include limitations. For somatic variants, limitations may include tumor purity, sample quality, and the absence of matched normal tissue. For germline variants, limitations may include the presence of VUS and the need for family segregation studies. ## Records and Measurements ### Documentation Requirements Laboratories must document the variant interpretation process for each reported variant. The documentation should include the framework used, the evidence criteria applied, the knowledge base versions, and the classification date. For germline variants, the documentation should include the ClinGen specification version when applicable. For somatic variants, the documentation should include the tier assignment and the clinical actionability evidence. The documentation should also include the variant calling parameters, quality metrics, and filtering criteria. This documentation supports reproducibility and allows the laboratory to review classification decisions when new evidence becomes available. ### Quality Metrics Laboratories should track quality metrics for variant interpretation. These metrics include the VUS rate for germline testing, the Tier III rate for somatic testing, and the inter-laboratory concordance rate. The ClinGen TP53 VCEP reported that their updated specifications decreased VUS rates and increased inter-laboratory concordance, demonstrating the value of tracking these metrics. The laboratory should also track the reclassification rate. Variants initially classified as VUS or Tier III may be reclassified as new evidence becomes available. The laboratory should have a process for reviewing and reclassifying variants when evidence changes. ### Audit and Review Processes The laboratory should conduct periodic audits of variant classifications. The audit should review a sample of classifications to ensure that the framework was applied correctly and that the evidence documentation is complete. The audit should also review the currency of the knowledge bases and specifications used. The audit findings should be documented and used to improve the variant interpretation process. Common findings include incomplete evidence documentation, outdated knowledge base versions, and inconsistent application of criteria. ## Common Failure Patterns ### Applying Germline Criteria to Somatic Variants A common failure is applying ACMG/AMP criteria to somatic variants. This failure produces misleading classifications because the evidence types differ. For example, population frequency data used in the ACMG/AMP framework may not be relevant for somatic variants, which are not present in the germline population. Similarly, segregation data used in the germline framework does not apply to somatic variants. The AMP/ASCO/CAP tier system was designed for somatic variants and should be used for tumor profiling. The tier system focuses on clinical actionability, which is the relevant question for cancer treatment decisions. ### Confusing Somatic Variants With Germline Variants Another common failure is confusing somatic variants with germline variants. This confusion occurs when the laboratory does not have matched normal tissue or does not adequately filter germline variants from somatic analysis. The CUP study demonstrated the importance of paired germline DNA analysis for distinguishing somatic, germline, and CHIP-related variants. When matched normal tissue is unavailable, the laboratory should use variant allele frequency thresholds and population databases to identify likely germline variants. Variants present at high allele frequency in the tumor sample and present in population databases are likely germline. Variants present at low allele frequency and absent from population databases are likely somatic. ### Ignoring Clonal Hematopoiesis Clonal hematopoiesis of indeterminate potential (CHIP) is a common source of false somatic variant calls. CHIP-associated variants are present in blood cells and can be detected in liquid biopsy samples. The CUP study identified likely CHIP-associated variants in 17 of 39 patients, defined as mutations in CHIP-associated genes with VAF at or above 2% in PBMC-derived genomic DNA. The ClinGen TP53 VCEP incorporated variant allele fraction as evidence of pathogenicity in the context of clonal hematopoiesis. This refinement acknowledges that CHIP-associated variants require different interpretation than true somatic variants. Laboratories performing liquid biopsy testing should have explicit criteria for identifying and reporting CHIP-associated variants. ### Overinterpreting Variants of Unknown Significance Overinterpreting VUS or Tier III variants is a common failure in both frameworks. In the germline framework, VUS should not be used for clinical decision-making. In the somatic framework, Tier III variants should not be used for treatment decisions. The laboratory should clearly communicate the uncertainty associated with these classifications. The ClinGen TP53 VCEP reported that many TP53 variants remain classified as VUS, and their updated specifications were designed to reduce VUS rates. This example demonstrates that VUS reduction is an ongoing challenge that requires gene-specific data and quantitative methods. ## Limitations and Professional Escalation Criteria ### Framework Limitations The ACMG/AMP framework has known limitations. The framework was designed for Mendelian disease genes and may not apply to genes with complex disease mechanisms. The framework relies on expert judgment for many criteria, which can lead to inter-laboratory variability. The ClinGen VCEP process addresses some of these limitations through gene-specific specifications and quantitative methods. The AMP/ASCO/CAP tier system also has limitations. The tier system depends on the currency of knowledge bases and professional guidelines. New therapies and evidence can change tier assignments. The tier system does not address all clinical questions, such as tumor mutational burden or microsatellite instability, which require separate interpretation frameworks. ### When to Escalate to Professional Consultation Laboratory professionals should escalate variant interpretation questions to professional consultation in several situations. These include variants with conflicting evidence, variants in genes with unclear disease mechanisms, variants with therapeutic implications that require expert review, and variants with potential germline implications detected in somatic testing. The CUP study identified pathogenic germline variants in MITF, NTRK1, and BAP1 in three patients. These germline findings have implications for the patients and their family members, even though the testing was performed for somatic variant analysis. The laboratory should have a process for identifying and reporting clinically significant germline findings detected in somatic testing. ### Safety and Regulatory Context Variant interpretation has direct implications for patient care. Incorrect classifications can lead to inappropriate treatment decisions, missed hereditary risk, or unnecessary surveillance. The laboratory should have a quality management system that addresses variant interpretation accuracy, documentation, and continuous improvement. The laboratory should also be aware of regulatory requirements for variant interpretation. These requirements may include proficiency testing, quality control, and reporting standards. The laboratory should document compliance with applicable regulations and professional guidelines. ## Bioinformatics Training and Reproducible Workflow Resources ### Foundational Training Pathways Laboratory professionals entering variant interpretation need structured training in bioinformatics methods. The European Bioinformatics Institute provides training pathways covering data resources, sequence analysis, and practical analysis education. These pathways help laboratory staff build the computational skills needed for variant calling and interpretation workflows. The Galaxy Training Network offers accessible workflow training and analysis tutorials that emphasize reproducibility. These tutorials cover variant calling pipelines, quality control, and annotation steps that are directly applicable to both germline and somatic analysis. The hands-on format allows laboratory professionals to practice workflows before implementing them in production. ### Reproducible Pipeline Standards Reproducible analysis pipelines are essential for clinical variant interpretation. The nf-core documentation describes community pipeline standards, usage, configuration, and reproducible workflow context. These standards help laboratories implement version-controlled pipelines that produce consistent results across runs and across laboratories. The Bioconductor project provides official package, workflow, installation, and reproducible genomic-analysis documentation. Many variant annotation and classification tools are available through Bioconductor, and the documentation supports proper installation and usage. Laboratories should verify that their analysis environment uses documented versions of these packages. ### Computing and Data Skills The Carpentries lessons provide foundational computing, data, shell, Git, and programming training context. These skills are necessary for laboratory professionals who manage variant calling pipelines, track analysis versions, and document workflow changes. Version control through Git is particularly important for maintaining reproducible analysis workflows. The National Center for Biotechnology Information provides official descriptions of NCBI databases, search systems, sequence resources, and analysis services. Laboratory professionals should understand how to use these resources for variant annotation, population frequency checks, and database cross-referencing. ## A Practical Decision Framework for Selecting Between Germline and Somatic Interpretation Systems ### The Core Decision Point: Sample Type and Clinical Question The first decision in any variant interpretation workflow is whether the testing context is germline, somatic, or mixed. This decision determines the classification framework, the evidence standards, and the reporting format. A laboratory that receives a tumor specimen for targeted therapy selection must use the AMP/ASCO/CAP tier system. A laboratory that receives a blood specimen for hereditary cancer risk assessment must use the ACMG/AMP framework with ClinGen specifications where available. A laboratory that receives paired tumor and normal specimens may need both frameworks, depending on the clinical question. The decision tree begins with three questions. First, what is the specimen type? Blood, saliva, and buccal samples are typically germline. Tumor tissue and liquid biopsy samples are typically somatic. Second, what is the clinical question? Hereditary risk, carrier status, and pharmacogenetic dosing require germline interpretation. Targeted therapy selection, resistance monitoring, and tumor classification require somatic interpretation. Third, does the test require paired analysis? Tumor-normal paired testing requires both frameworks because the normal sample provides the germline baseline and the tumor sample provides the somatic alterations. A practical example clarifies the decision process. A clinical molecular pathologist receives a liquid biopsy sample from a patient with cancer of unknown primary. The clinical question is whether genomic profiling can identify actionable mutations for targeted therapy. The specimen is somatic, the clinical question is therapeutic, and the appropriate framework is AMP/ASCO/CAP. The CUP study applied a 92-gene targeted sequencing panel to liquid biopsy samples and classified variants using the tier system, identifying 44 clinically relevant variants (Tier I-III) and actionable mutations in 13 genes. This example demonstrates that the somatic framework applies when the clinical question is tumor biology and treatment selection. ### The Mixed Testing Context: When Both Frameworks Apply Some testing contexts require both germline and somatic interpretation. Tumor-normal paired testing is the most common example. The normal sample is analyzed using the ACMG/AMP framework to identify germline variants that may contribute to hereditary cancer risk. The tumor sample is analyzed using the AMP/ASCO/CAP tier system to identify somatic variants that drive tumor growth and predict therapy response. The two analyses are complementary and must be reported separately. The CUP study illustrates the importance of paired analysis. The study analyzed circulating cell-free DNA together with paired germline DNA when available. Somatic, germline, and CHIP-related variants were classified using predefined variant allele frequency thresholds and paired ccfDNA/gDNA comparison. The study identified pathogenic germline variants in MITF, NTRK1, and BAP1 in three patients. These germline findings have implications for the patients and their family members, even though the testing was performed for somatic variant analysis. The NTRK1 germline variant was accompanied by an independent somatic mutation in the same gene, demonstrating that both frameworks may be needed for a single gene in a single patient. The laboratory must have a process for handling incidental germline findings in somatic testing. This process should include criteria for identifying likely germline variants, procedures for confirming germline origin, and a reporting pathway that includes genetic counseling recommendations. The laboratory should document this process in its standard operating procedures and train staff on the distinction between germline and somatic findings. ### A Structured Decision Matrix for Framework Selection The following decision matrix provides a structured approach to framework selection. The matrix uses four criteria: specimen type, clinical question, gene content, and reporting requirement. | Criterion | Germline Framework (ACMG/AMP with ClinGen) | Somatic Framework (AMP/ASCO/CAP) | |-----------|-------------------------------------------|----------------------------------| | Specimen type | Blood, saliva, buccal, fibroblasts | Tumor tissue, liquid biopsy, cell-free DNA | | Clinical question | Hereditary risk, carrier status, pharmacogenetics | Targeted therapy, resistance, diagnosis, prognosis | | Gene content | Mendelian disease genes with established disease mechanism | Cancer genes with therapeutic or diagnostic relevance | | Reporting requirement | Five-tier classification with evidence criteria | Tier I-IV classification with actionability evidence | | Matched normal required | No | Yes, for definitive somatic confirmation | | Population databases | gnomAD, ClinVar, locus-specific databases | COSMIC, OncoKB, CIViC, ClinVar | | ClinGen specifications | Apply where VCEP specifications exist | Not directly applicable, but VCEP data may inform somatic interpretation | The matrix is not a substitute for professional judgment. Some variants require case-by-case evaluation. For example, a TP53 variant detected in a tumor sample could be somatic, germline, or CHIP-associated. The interpretation depends on the variant allele fraction, the presence of the variant in matched normal tissue, and the clinical context. The ClinGen TP53 Variant Curation Expert Panel has updated its specifications to incorporate variant allele fraction as evidence of pathogenicity, particularly in the context of clonal hematopoiesis. This refinement acknowledges that the same gene and variant may require different interpretive frameworks depending on the tissue context. ### Implementing the Decision Framework in Laboratory Practice The decision framework should be implemented as a written standard operating procedure. The procedure should include the decision matrix, the criteria for each framework, and the documentation requirements. The procedure should be reviewed annually and updated when new ClinGen specifications or AMP/ASCO/CAP guidance becomes available. The implementation should include a pre-analytic review step. Before variant calling begins, the laboratory should document the specimen type, the clinical question, and the framework selection. This documentation should be part of the test requisition and should be reviewed by the laboratory director or designee. The pre-analytic review prevents framework misapplication and ensures that the appropriate evidence standards are applied. The implementation should also include a post-analytic review step. After variant classification, the laboratory should verify that the framework was applied correctly and that the evidence documentation is complete. This review should include a check of the ClinGen CSpec registry for the most current specification version when germline interpretation is performed. The ClinGen Criteria Specifications Registry provides the most current version of each VCEP specification, and the TP53 panel reported that updated specifications led to clinically meaningful classifications for 93% of pilot variants. ### Records and Measurements for Framework Selection The laboratory should maintain records that document framework selection for each test. These records should include the specimen type, the clinical question, the framework used, and the rationale for the selection. The records should also include the version of the framework or specification used, the date of the analysis, and the personnel who performed the interpretation. The laboratory should track metrics that measure the effectiveness of framework selection. These metrics include the rate of germline findings in somatic testing, the rate of somatic findings in germline testing, and the rate of reclassification due to framework misapplication. The CUP study identified pathogenic germline variants in 3 of 39 patients, demonstrating that germline findings in somatic testing are not rare. The laboratory should monitor this rate and review cases where germline findings were missed or misclassified. The laboratory should also track the VUS rate for germline testing and the Tier III rate for somatic testing. The ClinGen TP53 VCEP reported that updated specifications decreased VUS rates and increased inter-laboratory concordance. The laboratory should compare its VUS and Tier III rates with published benchmarks and investigate outliers. ### Common Failure Patterns in Framework Selection The most common failure is applying the germline framework to somatic variants. This failure occurs when the laboratory does not recognize that the testing context is somatic or when the laboratory uses a single interpretation pipeline for all variants. The result is misleading classifications because the evidence types differ. Population frequency data used in the ACMG/AMP framework may not be relevant for somatic variants, and segregation data does not apply to somatic variants. The second most common failure is confusing somatic variants with germline variants. This confusion occurs when the laboratory does not have matched normal tissue or does not adequately filter germline variants from somatic analysis. The CUP study demonstrated the importance of paired germline DNA analysis for distinguishing somatic, germline, and CHIP-related variants. When matched normal tissue is unavailable, the laboratory should use variant allele frequency thresholds and population databases to identify likely germline variants. The third common failure is ignoring clonal hematopoiesis. CHIP-associated variants are present in blood cells and can be detected in liquid biopsy samples. The CUP study identified likely CHIP-associated variants in 17 of 39 patients, defined as mutations in CHIP-associated genes with VAF at or above 2% in PBMC-derived genomic DNA. Laboratories performing liquid biopsy testing should have explicit criteria for identifying and reporting CHIP-associated variants. ### Professional Escalation Criteria The laboratory should escalate framework selection questions to professional consultation in several situations. These include cases where the specimen type is ambiguous, cases where the clinical question spans both germline and somatic contexts, and cases where the variant has potential germline implications detected in somatic testing. The CUP study identified pathogenic germline variants in MITF, NTRK1, and BAP1, and the NTRK1 germline variant was accompanied by an independent somatic mutation in the same gene. These cases require expert review to ensure that both germline and somatic implications are reported appropriately. The laboratory should also escalate cases where the variant classification has direct therapeutic implications that require expert review. This includes variants in genes with emerging evidence for targeted therapy, variants with conflicting evidence across knowledge bases, and variants in genes with unclear disease mechanisms. The escalation process should be documented and should include a defined timeline for review. ## Frequently Asked Questions ### What is the primary difference between ACMG/AMP and AMP/ASCO/CAP frameworks? The ACMG/AMP framework classifies germline variants for hereditary disease risk using 28 evidence criteria that weigh population frequency, segregation, functional studies, and disease mechanism. The AMP/ASCO/CAP framework classifies somatic variants in cancer using four tiers based on clinical actionability, including therapeutic targets, resistance mechanisms, and diagnostic or prognostic significance. The choice of framework depends on whether the testing question is hereditary risk or tumor biology. ### When should a laboratory use ClinGen specifications instead of generic ACMG/AMP criteria? A laboratory should use ClinGen specifications when a Variant Curation Expert Panel has published gene-specific criteria for the gene being tested. The ClinGen Criteria Specifications Registry provides the most current version of each specification. The ClinGen TP53 Expert Panel demonstrated that gene-specific specifications reduce VUS rates and increase inter-laboratory concordance compared with generic criteria. ### How does variant allele frequency inform somatic variant classification? Variant allele frequency helps distinguish somatic variants from germline variants and CHIP-associated variants. A variant present at high allele frequency in tumor tissue but absent in matched normal tissue is likely somatic. A variant present at similar frequency in both tissues may be germline. The CUP study used predefined VAF thresholds and paired ccfDNA/gDNA comparison to classify variants, and the ClinGen TP53 panel incorporated VAF as evidence of pathogenicity in the context of clonal hematopoiesis. ### What is the role of matched normal tissue in somatic variant interpretation? Matched normal tissue is essential for distinguishing somatic variants from germline variants. Without matched normal tissue, the laboratory must rely on VAF thresholds, population databases, and gene-specific knowledge to make this distinction. The CUP study demonstrated that paired germline DNA analysis is critical for identifying pathogenic germline variants that may be missed in somatic-only analysis. ### How should a laboratory report variants of uncertain significance? Variants of uncertain significance in the germline framework and Tier III variants in the somatic framework should be reported with clear language indicating that clinical significance is unknown. The report should not suggest clinical action based on VUS or Tier III variants. The laboratory should have a process for reclassifying variants as new evidence becomes available. ### What is clonal hematopoiesis and why does it matter for somatic variant interpretation? Clonal hematopoiesis of indeterminate potential refers to somatic mutations in blood cells that are not associated with hematologic malignancy but can be detected in liquid biopsy samples. The CUP study identified CHIP-associated variants in 17 of 39 patients. CHIP-associated variants can be mistaken for tumor-derived somatic variants, leading to incorrect treatment decisions. Laboratories should have explicit criteria for identifying and reporting CHIP-associated variants. ### How do ClinGen API microservices support variant classification workflows? ClinGen API microservices expose findable, accessible, interoperable, reusable, and AI-ready variant data. These services allow laboratories to integrate ClinGen specifications and variant data into local interpretation pipelines. The API infrastructure supports next-generation software applications and AI systems for variant classification. ### What should a laboratory do when a somatic test reveals a pathogenic germline variant? The laboratory should have a process for identifying and reporting clinically significant germline findings detected in somatic testing. The CUP study identified pathogenic germline variants in MITF, NTRK1, and BAP1 in three patients. These findings have implications for the patients and their family members. The laboratory should report these findings with appropriate context and recommend genetic counseling. ## Related Bioinformatics Guides - [Digital Pathology Validation: A Practical Guide to CAP and RCPath Compliance](/knowledge/bioinformatics/digital-pathology-validation-a-practical-guide-to-cap-and-rcpath-compliance) - [FAIR Data Maturity Model: A Practical Assessment Framework for Bioinformatics Workflows](/knowledge/bioinformatics/fair-data-maturity-model-a-practical-assessment-framework-for-bioinformatics-workflows) - [Detecting Structural Variants with Long-Read Sequencing: Methods and Considerations](/knowledge/bioinformatics/detecting-structural-variants-with-long-read-sequencing-methods-and-considerations) - [RNA-Seq vs qPCR: Validation and Comparison](/knowledge/bioinformatics/rna-seq-vs-qpcr-validation-and-comparison) - [Digital Pathology Guidelines: A Reference for Implementation](/knowledge/bioinformatics/digital-pathology-guidelines-a-reference-for-implementation) ## Related Clinical & Scientific Guides * [A Practical Guide to Detecting Antimicrobial Resistance Genes in Shotgun Metagenomic Data](/knowledge/bioinformatics/a-practical-guide-to-detecting-antimicrobial-resistance-genes-in-shotgun-metagenomic-data) * [Computational Immunology: Modeling the Immune System](/knowledge/bioinformatics/computational-immunology-modeling-the-immune-system) * [How to Set Hard Filters for Germline Variant Calling: A Practical Guide to GATK Best Practices](/knowledge/bioinformatics/how-to-set-hard-filters-for-germline-variant-calling-a-practical-guide-to-gatk-best-practices) ## References and Further Reading - [NCBI Data Resources](https://www.ncbi.nlm.nih.gov/). National Center for Biotechnology Information. - [EMBL-EBI Training](https://www.ebi.ac.uk/training). European Bioinformatics Institute. - [Bioconductor](https://bioconductor.org/). Bioconductor Project. - [Galaxy Training Network](https://training.galaxyproject.org/). Galaxy Project. - [nf-core Documentation](https://nf-co.re/docs). nf-core. - [The Carpentries Lessons](https://carpentries.org/lessons). The Carpentries. - [ClinGen API platform for classification of human genetic variants.](https://doi.org/10.1016/j.xgen.2026.101211). 2026. - [Cancer of unknown primary genomic profiling from cell-free DNA provides insights into CUP biology and vulnerabilities.](https://doi.org/10.3389/fmed.2026.1777582). 2026. - [A quantitative, Bayesian-informed approach to gene-specific variant classification: Updated Expert Panel recommendations improve classification of TP53 germline variants for Li-Fraumeni syndrome.](https://doi.org/10.1186/s13073-025-01536-3). 2025. - [Somatic mosaicism in the δ-aminolevulinate dehydratase gene causing late-onset porphyria with erythroid-driven pathogenesis.](https://doi.org/10.1016/j.ymgmr.2026.101320). 2026. > This article is educational and does not replace validated analysis plans, institutional policy, clinical interpretation, or specialist review.