CIViC (www.civicdb.org) is a free and open knowledgebase that leverages public curation together with expert moderation to address the bottleneck created with the need to interpret large numbers of variants from next generation sequencing of tumor DNA. Curation from literature and meeting abstracts is utilized to create Evidence Items (EIDs), and collections of EIDs are summarized into Assertions which reflect the state of the field for a variant. Assertions incorporate variant classification standards such as those from AMP/ASCO/CAP for clinical actionability, or ClinGen/CGC/VICC guidelines for oncogenicity. The CIViC data model has consistently developed to better capture the considerable variation of cancer. CIViC employs a flexible model for gene variants which may be combined into multi-gene Molecular Profiles. While gene mutations play the largest role in personalized medicine, other entities such as tumor mutation burden (TMB), microsatellite instability (MSI) or homologous recombination repair deficiency (HRD) are increasingly used as diagnostic, prognostic or therapeutic markers. To capture entities such as these, CIViC has introduced a new feature type, which models tumorbiomarkers not directly associated with genes or specific regions of the genome. This new type of biomarker accompanies Genes as a formal Feature entity in the CIViC data model. Like Genes, each of these biomarkers will have a page in CIViC, and be associated with an NCI thesaurus entry whenever possible. CIViC is developing a generalized Feature-Variant data model, enabling the addition of new Feature types in future updates, such as large genomic regions.
{"title":"28. Addition of non-gene features to the CIViC data model","authors":"Arpad Danos , Kilannin Krysiak , Adam Coffman , Joshua McMichael , Mariam Khanfar , Cameron Grisdale , Alex Wagner , Malachi Griffith , Obi Griffith","doi":"10.1016/j.cancergen.2024.08.030","DOIUrl":"10.1016/j.cancergen.2024.08.030","url":null,"abstract":"<div><div>CIViC (<span><span>www.civicdb.org</span><svg><path></path></svg></span>) is a free and open knowledgebase that leverages public curation together with expert moderation to address the bottleneck created with the need to interpret large numbers of variants from next generation sequencing of tumor DNA. Curation from literature and meeting abstracts is utilized to create Evidence Items (EIDs), and collections of EIDs are summarized into Assertions which reflect the state of the field for a variant. Assertions incorporate variant classification standards such as those from AMP/ASCO/CAP for clinical actionability, or ClinGen/CGC/VICC guidelines for oncogenicity. The CIViC data model has consistently developed to better capture the considerable variation of cancer. CIViC employs a flexible model for gene variants which may be combined into multi-gene Molecular Profiles. While gene mutations play the largest role in personalized medicine, other entities such as tumor mutation burden (TMB), microsatellite instability (MSI) or homologous recombination repair deficiency (HRD) are increasingly used as diagnostic, prognostic or therapeutic markers. To capture entities such as these, CIViC has introduced a new feature type, which models tumorbiomarkers not directly associated with genes or specific regions of the genome. This new type of biomarker accompanies Genes as a formal Feature entity in the CIViC data model. Like Genes, each of these biomarkers will have a page in CIViC, and be associated with an NCI thesaurus entry whenever possible. CIViC is developing a generalized Feature-Variant data model, enabling the addition of new Feature types in future updates, such as large genomic regions.</div></div>","PeriodicalId":49225,"journal":{"name":"Cancer Genetics","volume":"286 ","pages":"Page S9"},"PeriodicalIF":1.4,"publicationDate":"2024-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142323393","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-01DOI: 10.1016/j.cancergen.2024.08.066
Durga Prasad Dash , Eric Severson , Kyle Strickland , Heidi Ko , Rebeccaann Previs , Stephanie Hastings , Michelle Green , Paul Depietro , Brian Caveney , Marcia Eisenberg , Taylor Jensen , Jeffrey Conroy , Shakti Ramkissoon , Shengle Zhang
Background
Colorectal cancer (CRC) ranks third in terms of new tumor cases and the second leading cause of cancer-related death worldwide (PMID: 30207593). KRAS is one of the most frequently mutated oncogenes in CRC, with approximately 40% of patients harboring activating missense mutations in KRAS (PMID: 31972237). Patients with KRAS-mutant CRC have a worse prognosis than those with KRAS wild-type CRC [PMID: 20008640; PMID: 28453697). Here we report a rare finding of three clinically significant KRAS mutations co-occurring in a patient with microsatellite stable colorectal adenocarcinoma.
Methods
Comprehensive genomic and immune profiling (CGIP) was performed on a hemicolectomy specimen from a >80 year old patient with advanced colorectal adenocarcinoma with 60% tumor nuclei and more than 1000 neoplastic cells per slide at a CAP/CLIA and NYS CLEP certified reference laboratory with the OmniSeq® INSIGHT test (PMID: 34855780). OmniSeq INSIGHT is a next generation sequencing-based laboratory developed test for both DNA and RNA for the detection of genomic and transcriptomic variants, in formalin-fixed paraffin-embedded (FFPE) tumor tissue.
Results
We identified three co-occurring KRAS mutations (c.35G>A p.G12D, c.38G>A p.G13D and c.351A>T p.K117N) with VAF 10.1%, 9.5% and 4.1% respectively in the same colorectal adenocarcinoma patient specimen. All three mutations are associated with resistance to targeted therapies with cetuximab and panitumumab. In addition, the sequencing utilized was able to reveal that G12D and G13D mutations occurred in different cell clones/populations.
Conclusions
CGIP revealed three distinct KRAS co-mutations at known KRAS hot-spots. In addition, CGIP can distinguish allele-specific KRAS mutations and tumoral sub-clonal populations.
{"title":"64. A rare finding of triple KRAS mutations with OmniSeq® INSIGHT in a patient with colorectal adenocarcinoma","authors":"Durga Prasad Dash , Eric Severson , Kyle Strickland , Heidi Ko , Rebeccaann Previs , Stephanie Hastings , Michelle Green , Paul Depietro , Brian Caveney , Marcia Eisenberg , Taylor Jensen , Jeffrey Conroy , Shakti Ramkissoon , Shengle Zhang","doi":"10.1016/j.cancergen.2024.08.066","DOIUrl":"10.1016/j.cancergen.2024.08.066","url":null,"abstract":"<div><h3>Background</h3><div>Colorectal cancer (CRC) ranks third in terms of new tumor cases and the second leading cause of cancer-related death worldwide (PMID: 30207593). <em>KRAS</em> is one of the most frequently mutated oncogenes in CRC, with approximately 40% of patients harboring activating missense mutations in <em>KRAS</em> (PMID: 31972237). Patients with <em>KRAS</em>-mutant CRC have a worse prognosis than those with <em>KRAS</em> wild-type CRC [PMID: 20008640; PMID: 28453697). Here we report a rare finding of three clinically significant <em>KRAS</em> mutations co-occurring in a patient with microsatellite stable colorectal adenocarcinoma.</div></div><div><h3>Methods</h3><div>Comprehensive genomic and immune profiling (CGIP) was performed on a hemicolectomy specimen from a >80 year old patient with advanced colorectal adenocarcinoma with 60% tumor nuclei and more than 1000 neoplastic cells per slide at a CAP/CLIA and NYS CLEP certified reference laboratory with the OmniSeq® INSIGHT test (PMID: 34855780). OmniSeq INSIGHT is a next generation sequencing-based laboratory developed test for both DNA and RNA for the detection of genomic and transcriptomic variants, in formalin-fixed paraffin-embedded (FFPE) tumor tissue.</div></div><div><h3>Results</h3><div>We identified three co-occurring <em>KRAS</em> mutations (c.35G>A p.G12D, c.38G>A p.G13D and c.351A>T p.K117N) with VAF 10.1%, 9.5% and 4.1% respectively in the same colorectal adenocarcinoma patient specimen. All three mutations are associated with resistance to targeted therapies with cetuximab and panitumumab. In addition, the sequencing utilized was able to reveal that G12D and G13D mutations occurred in different cell clones/populations.</div></div><div><h3>Conclusions</h3><div>CGIP revealed three distinct <em>KRAS</em> co-mutations at known KRAS hot-spots. In addition, CGIP can distinguish allele-specific <em>KRAS</em> mutations and tumoral sub-clonal populations.</div></div>","PeriodicalId":49225,"journal":{"name":"Cancer Genetics","volume":"286 ","pages":"Page S21"},"PeriodicalIF":1.4,"publicationDate":"2024-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142323398","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-01DOI: 10.1016/j.cancergen.2024.08.035
Rachel Karchin , Jasmine Baker , Kyle Moad , Kyle Anderson , Madison Larsen , Supra Gajjala , James Higgins , Ben Busby
OpenCRAVAT is an open-source modular variant meta-annotator designed to make variant interpretation accessible to a wide audience. The modular design allows researchers to design customized workflows and utilize a diverse set of analysis methods, fostering a personalized approach to variant interpretation. While valuable information about variants is scattered across hundreds of databases and computational variant effect predictors, OpenCRAVAT centralizes access to these tools and makes them available through an easy-to-use interface. Hundreds of tools can be installed using point-and-click or simple command-line statements, and results are combined and presented in a variety of output formats. These tools cater to a broad range of variant types, encompassing germline, somatic, common, rare, coding and non-coding variants. We have recently introduced 'Packages', a new feature that consists of pre-configured combinations of annotators, filters, and output layouts that are focused on specific variant-related questions. For example, the Drug Interaction Package is designed to find pharmacogenomic annotations; the Hereditary Cancer Package finds rare variants in known hereditary cancer genes; and the Splicing Package focuses on variants associated with aberrant splicing. We are now adding support for annotation of large structural variants, comparison of genetic variants across user-defined patient cohorts, and a workflow for annotation of very large (UK-Biobank sized) cohorts in the cloud. To support neoantigen discovery, we provide annotations of epitopes in the Cancer Epitope Database and Analysis Resource (CEDAR). Finally, we provide a Single Variant Report application customized to the needs of molecular tumor boards, to assist oncologists in making treatment decisions.
{"title":"33. Customize your variant interpretation workflow with OpenCRAVAT","authors":"Rachel Karchin , Jasmine Baker , Kyle Moad , Kyle Anderson , Madison Larsen , Supra Gajjala , James Higgins , Ben Busby","doi":"10.1016/j.cancergen.2024.08.035","DOIUrl":"10.1016/j.cancergen.2024.08.035","url":null,"abstract":"<div><div>OpenCRAVAT is an open-source modular variant meta-annotator designed to make variant interpretation accessible to a wide audience. The modular design allows researchers to design customized workflows and utilize a diverse set of analysis methods, fostering a personalized approach to variant interpretation. While valuable information about variants is scattered across hundreds of databases and computational variant effect predictors, OpenCRAVAT centralizes access to these tools and makes them available through an easy-to-use interface. Hundreds of tools can be installed using point-and-click or simple command-line statements, and results are combined and presented in a variety of output formats. These tools cater to a broad range of variant types, encompassing germline, somatic, common, rare, coding and non-coding variants. We have recently introduced 'Packages', a new feature that consists of pre-configured combinations of annotators, filters, and output layouts that are focused on specific variant-related questions. For example, the Drug Interaction Package is designed to find pharmacogenomic annotations; the Hereditary Cancer Package finds rare variants in known hereditary cancer genes; and the Splicing Package focuses on variants associated with aberrant splicing. We are now adding support for annotation of large structural variants, comparison of genetic variants across user-defined patient cohorts, and a workflow for annotation of very large (UK-Biobank sized) cohorts in the cloud. To support neoantigen discovery, we provide annotations of epitopes in the Cancer Epitope Database and Analysis Resource (CEDAR). Finally, we provide a Single Variant Report application customized to the needs of molecular tumor boards, to assist oncologists in making treatment decisions.</div></div>","PeriodicalId":49225,"journal":{"name":"Cancer Genetics","volume":"286 ","pages":"Page S11"},"PeriodicalIF":1.4,"publicationDate":"2024-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142323287","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Publicly available artificial intelligence (AI) chatbot-style search engines are gaining popularity for various applications, ranging from writing poetry to determining oncogenic cancer variants and genes. However, to our knowledge, a systematic evaluation of the effectiveness of these tools in cancer variant interpretation is lacking in current literature.
In this proof-of-concept study, four free online AI-assisted search engines (ChatGPT, Perplexity AI, Claude AI, and Llama2) were given simple standardized queries to investigate the clinical relevance of multiple gene variants observed in lung adenocarcinoma, glioma, and acute myeloid leukemia. The queries were structured as follows: 'What is the clinical significance of [GENE] [protein-level (p.) nomenclature] in [cancer type]?'
As anticipated, variants of uncertain significance (VUS) illustrated challenges for using AI-assisted search engines in cancer variant interpretation. Each tool incorrectly attributed oncogenicity to at least 1 of the 6 VUS investigated: Perplexity AI (1/6 VUS incorrectly represented as oncogenic), ChatGPT (2/6), Llama2 (4/6), Claude AI (5/6).
The overestimation of oncogenicity in these tools may be driven by conditioning of these AI-assisted search engines by past and current users for positive assignation attributes or from application of a response format with incorrect extrapolation of studies describing variants in the same gene without the ability to draw nuanced conclusions from studies focusing on different aspects of gene function. While there are challenges in using AI-assisted search engines in the clinical genomic space currently, this rapidly improving technology could provide a useful supplement for cancer variant analysts when combined with caution and expert human oversight.
{"title":"63. An introduction to publicly available AI-assisted chatbot-style search engines for cancer variant curation","authors":"Beth Pitel, Antonina Wojcik, Christy Koellner, Claire Teigen, Katherine Geiersbach, Patricia Greipp, Xinjie Xu, Cinthya Zepeda Mendoza","doi":"10.1016/j.cancergen.2024.08.065","DOIUrl":"10.1016/j.cancergen.2024.08.065","url":null,"abstract":"<div><div>Publicly available artificial intelligence (AI) chatbot-style search engines are gaining popularity for various applications, ranging from writing poetry to determining oncogenic cancer variants and genes. However, to our knowledge, a systematic evaluation of the effectiveness of these tools in cancer variant interpretation is lacking in current literature.</div><div>In this proof-of-concept study, four free online AI-assisted search engines (ChatGPT, Perplexity AI, Claude AI, and Llama2) were given simple standardized queries to investigate the clinical relevance of multiple gene variants observed in lung adenocarcinoma, glioma, and acute myeloid leukemia. The queries were structured as follows: 'What is the clinical significance of [GENE] [protein-level (p.) nomenclature] in [cancer type]?'</div><div>As anticipated, variants of uncertain significance (VUS) illustrated challenges for using AI-assisted search engines in cancer variant interpretation. Each tool incorrectly attributed oncogenicity to at least 1 of the 6 VUS investigated: Perplexity AI (1/6 VUS incorrectly represented as oncogenic), ChatGPT (2/6), Llama2 (4/6), Claude AI (5/6).</div><div>The overestimation of oncogenicity in these tools may be driven by conditioning of these AI-assisted search engines by past and current users for positive assignation attributes or from application of a response format with incorrect extrapolation of studies describing variants in the same gene without the ability to draw nuanced conclusions from studies focusing on different aspects of gene function. While there are challenges in using AI-assisted search engines in the clinical genomic space currently, this rapidly improving technology could provide a useful supplement for cancer variant analysts when combined with caution and expert human oversight.</div></div>","PeriodicalId":49225,"journal":{"name":"Cancer Genetics","volume":"286 ","pages":"Page S20"},"PeriodicalIF":1.4,"publicationDate":"2024-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142323399","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-01DOI: 10.1016/j.cancergen.2024.08.075
Digvijay Yadav, Shrey Sukhadia
Pathology relies on examining H&E-stained FFPE tissue sections via microscopy for diagnosis. Pathomics quantifies features from digitized FFPE images, or whole slide images (WSIs), reflecting tissue and cellular structures potentially linked to gene expression patterns seen in RNA sequencing. Although Deep Learning (DL) methods have advanced gene expression prediction from WSIs, understanding the pathomic features affecting model predictions is challenging.
Our study analyzed 89 FFPE breast cancer tissue slide images from the TCGA registry, extracting around 300 pathomic features using HistomicsTK. These features, representing cell morphometry, intensity, and gradient, were assessed within tumor regions annotated by pathologists. We selected the most heterogeneous and correlating features for a multitask regression model, predicting gene expression levels with high accuracy (AUC > 0.8) for two biomarkers, MFAP5 and MXRA8.
In contrast, the ResNet-50 DL model trained on random WSI patches showed lower AUC scores for these biomarkers and did not interpret pathomic features that contribute to gene expression predictions. Literature suggests MFAP5 upregulation in breast carcinomas correlates with poor prognosis, while MXRA8 modulates triple-negative breast cancer progression.
The study concludes that Pathomics-based Machine Learning outperforms DL in predicting gene expression from FFPE WSIs in invasive breast carcinoma, providing a more effective tool for understanding the disease at the molecular level.
{"title":"73. Pathomics-based machine mearning versus deep learning: Which is a better approach for whole slide image analyses?","authors":"Digvijay Yadav, Shrey Sukhadia","doi":"10.1016/j.cancergen.2024.08.075","DOIUrl":"10.1016/j.cancergen.2024.08.075","url":null,"abstract":"<div><div>Pathology relies on examining H&E-stained FFPE tissue sections via microscopy for diagnosis. Pathomics quantifies features from digitized FFPE images, or whole slide images (WSIs), reflecting tissue and cellular structures potentially linked to gene expression patterns seen in RNA sequencing. Although Deep Learning (DL) methods have advanced gene expression prediction from WSIs, understanding the pathomic features affecting model predictions is challenging.</div><div>Our study analyzed 89 FFPE breast cancer tissue slide images from the TCGA registry, extracting around 300 pathomic features using HistomicsTK. These features, representing cell morphometry, intensity, and gradient, were assessed within tumor regions annotated by pathologists. We selected the most heterogeneous and correlating features for a multitask regression model, predicting gene expression levels with high accuracy (AUC > 0.8) for two biomarkers, <em>MFAP5</em> and <em>MXRA8</em>.</div><div>In contrast, the ResNet-50 DL model trained on random WSI patches showed lower AUC scores for these biomarkers and did not interpret pathomic features that contribute to gene expression predictions. Literature suggests <em>MFAP5</em> upregulation in breast carcinomas correlates with poor prognosis, while <em>MXRA8</em> modulates triple-negative breast cancer progression.</div><div>The study concludes that Pathomics-based Machine Learning outperforms DL in predicting gene expression from FFPE WSIs in invasive breast carcinoma, providing a more effective tool for understanding the disease at the molecular level.</div></div>","PeriodicalId":49225,"journal":{"name":"Cancer Genetics","volume":"286 ","pages":"Page S23"},"PeriodicalIF":1.4,"publicationDate":"2024-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142323332","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-01DOI: 10.1016/j.cancergen.2024.08.052
Matthew Croken, Olga Lukatskaya
Comprehensive and whole exome NGS panels can generate large amounts of genomic information for a single specimen. This makes these approaches very powerful, however converting raw signals from the NGS instrument into actionable, clinical information requires specialized data pipelines. The expense and expertise required to deploy these pipelines may place NGS testing out of reach for clinics and research groups in low resource settings. TAGVar (Tertiary Analysis of Genomic Variants) is a freely available software tool that facilitates somatic variant classification in both the clinical and research contexts. The application takes VCF formatted genomic variants and outputs HGVS annotations, predicted effect, and links to external variant databases, like dbSNP, gnomAD, ClinVar, COSMIC, and CancerHotspots.org. TAGVar sorts and categorizes variants based on read coverage, variant allele frequency, inferred transcript effect, description in somatic variant databases, or presence in known cancer-related genes as well as additional user-defined criteria. In the clinic, these classifications streamline the identification of reportable variants. In research, the same classification scheme identifies known and novel somatic variants associated with disease. TAGVar has relatively low memory and CPU requirements and does not require a stable internet connection to run. These design features make TAGVar ideal for use in low resource settings.
{"title":"50. TAGVar: A simple, free software tool to annotate genomic variants for clinical review","authors":"Matthew Croken, Olga Lukatskaya","doi":"10.1016/j.cancergen.2024.08.052","DOIUrl":"10.1016/j.cancergen.2024.08.052","url":null,"abstract":"<div><div>Comprehensive and whole exome NGS panels can generate large amounts of genomic information for a single specimen. This makes these approaches very powerful, however converting raw signals from the NGS instrument into actionable, clinical information requires specialized data pipelines. The expense and expertise required to deploy these pipelines may place NGS testing out of reach for clinics and research groups in low resource settings. TAGVar (Tertiary Analysis of Genomic Variants) is a freely available software tool that facilitates somatic variant classification in both the clinical and research contexts. The application takes VCF formatted genomic variants and outputs HGVS annotations, predicted effect, and links to external variant databases, like dbSNP, gnomAD, ClinVar, COSMIC, and CancerHotspots.org. TAGVar sorts and categorizes variants based on read coverage, variant allele frequency, inferred transcript effect, description in somatic variant databases, or presence in known cancer-related genes as well as additional user-defined criteria. In the clinic, these classifications streamline the identification of reportable variants. In research, the same classification scheme identifies known and novel somatic variants associated with disease. TAGVar has relatively low memory and CPU requirements and does not require a stable internet connection to run. These design features make TAGVar ideal for use in low resource settings.</div></div>","PeriodicalId":49225,"journal":{"name":"Cancer Genetics","volume":"286 ","pages":"Page S16"},"PeriodicalIF":1.4,"publicationDate":"2024-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142323210","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-01DOI: 10.1016/j.cancergen.2024.08.056
Jason Saliba , Arpad Danos , Kilannin Krysiak , Adam Coffman , Susanna Kiwala , Joshua McMichael , Cameron J. Grisdale , Ian King , Shamini Selvarajah , Xinjie Xu , Rashmi Kanagal-Shamanna , Laveniya Satgunaseelan , David Meredith , Madina Sukhanova , Alanna J. Church , Larissa V. Furtado , Charles G. Mullighan , Peter Horak , Dmitriy Sonkin , Marco Tartaglia , Malachi Griffith
Interpretation of the clinical significance of somatic variants in cancer remains a major challenge for cancer diagnosis, prognosis, and predicting response to targeted therapies. The Clinical Genome Resource (ClinGen) has established tools, web resources and procedures to help communities of experts establish the clinical relevance of genes and variants. However, ClinGen's effort is almost exclusively focused on the interpretation of germline variants and their role in heritable phenotypes, leaving a significant gap in clinical interpretation of somatic variants in cancer. To address this need, the ClinGen Somatic Clinical Domain Working Group is creating a knowledgebase of high-quality assertions of the clinical significance of somatic variants in cancer within the CIViC platform to capture expert panel curation efforts and adapts the procedures of ClinGen germline groups to somatic variant interpretation. This effort will broadly enable research and clinical translation involving the use of somatic cancer variant knowledge. We have established processes to engage an expert community and facilitated the creation of eight Somatic Cancer Variant Curation Expert Panels (SC-VCEPs) with strategies to foster necessary expansion. Formation of these SC-VCEPs supports the creation of a ClinGen Somatic Knowledgebase of clinical cancer variant assertions curated and approved by experts. We are working to adopt and guide ongoing development of several emerging GA4GH standards that enable the Findable, Accessible, Interoperable, and Reusable (FAIR) principles for genomic knowledge sharing. Finally, we are using natural language processing approaches to accelerate a set of defined human knowledge curation tasks that currently limit the rate of expert curation.
{"title":"54. Creation of a knowledgebase of high-quality assertions of the clinical actionability of somatic variants in cancer","authors":"Jason Saliba , Arpad Danos , Kilannin Krysiak , Adam Coffman , Susanna Kiwala , Joshua McMichael , Cameron J. Grisdale , Ian King , Shamini Selvarajah , Xinjie Xu , Rashmi Kanagal-Shamanna , Laveniya Satgunaseelan , David Meredith , Madina Sukhanova , Alanna J. Church , Larissa V. Furtado , Charles G. Mullighan , Peter Horak , Dmitriy Sonkin , Marco Tartaglia , Malachi Griffith","doi":"10.1016/j.cancergen.2024.08.056","DOIUrl":"10.1016/j.cancergen.2024.08.056","url":null,"abstract":"<div><div>Interpretation of the clinical significance of somatic variants in cancer remains a major challenge for cancer diagnosis, prognosis, and predicting response to targeted therapies. The Clinical Genome Resource (ClinGen) has established tools, web resources and procedures to help communities of experts establish the clinical relevance of genes and variants. However, ClinGen's effort is almost exclusively focused on the interpretation of germline variants and their role in heritable phenotypes, leaving a significant gap in clinical interpretation of somatic variants in cancer. To address this need, the ClinGen Somatic Clinical Domain Working Group is creating a knowledgebase of high-quality assertions of the clinical significance of somatic variants in cancer within the CIViC platform to capture expert panel curation efforts and adapts the procedures of ClinGen germline groups to somatic variant interpretation. This effort will broadly enable research and clinical translation involving the use of somatic cancer variant knowledge. We have established processes to engage an expert community and facilitated the creation of eight Somatic Cancer Variant Curation Expert Panels (SC-VCEPs) with strategies to foster necessary expansion. Formation of these SC-VCEPs supports the creation of a ClinGen Somatic Knowledgebase of clinical cancer variant assertions curated and approved by experts. We are working to adopt and guide ongoing development of several emerging GA4GH standards that enable the Findable, Accessible, Interoperable, and Reusable (FAIR) principles for genomic knowledge sharing. Finally, we are using natural language processing approaches to accelerate a set of defined human knowledge curation tasks that currently limit the rate of expert curation.</div></div>","PeriodicalId":49225,"journal":{"name":"Cancer Genetics","volume":"286 ","pages":"Pages S17-S18"},"PeriodicalIF":1.4,"publicationDate":"2024-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142322858","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-01DOI: 10.1016/j.cancergen.2024.08.005
Eric McGinnis, Zeid Hamadeh, Tara Spence
Introduction
Optical genome mapping using the Bionano Saphyr platform has been deployed as the front-line diagnostic test for acute leukemias at Vancouver General Hospital. The new OGM platform iteration, Stratys, incorporates updates aimed at increasing throughput and is under evaluation to facilitate ramp-up of OGM-based testing in our clinical laboratory. We describe preliminary results of a clinical verification of Stratys, through comparison with the Saphyr, in diagnostic evaluation of hematologic malignancies.
Methods
We processed and annotated 10 specimens from patients with AML (6), ALL (3), and MDS (1) using manufacturer-recommended protocols on Saphyr (rare variant analysis pipeline) and Stratys (low allele fraction guided assembly pipeline). Data quality metrics and identified variants (filtered using provided control datasets and laboratory-developed gene-based filtering approaches) were compared between platforms.
Results
All 32 variants meeting our laboratory's clinical reporting standards using Saphyr were detected using Stratys. Data quality metrics differed negligibly for cases between platforms. Unfiltered Stratys data included substantially more duplication calls (median 64% increase over Saphyr) but population-filtered variants were similar in number between platforms. In representative variants (including translocations/deletions/chromosome gain/loss), breakpoint nucleotide positions were typically identical or, infrequently, differed by less than 10kb interplatform, and variant frequency measures typically differed by 8% on average (without a clear interplatform bias).
Conclusion
Preliminary analysis of verification data indicates Stratys to perform comparably to Saphyr for detection of reportable somatic variants in clinical bone marrow specimens, albeit with relative changes in variant call rates (for variants we currently consider non-reportable) and reported variant burdens.
{"title":"3. Interplatform comparison of Stratys and Saphyr: Preliminary results of OGM clinical verification in hematologic cancers","authors":"Eric McGinnis, Zeid Hamadeh, Tara Spence","doi":"10.1016/j.cancergen.2024.08.005","DOIUrl":"10.1016/j.cancergen.2024.08.005","url":null,"abstract":"<div><h3>Introduction</h3><div>Optical genome mapping using the Bionano Saphyr platform has been deployed as the front-line diagnostic test for acute leukemias at Vancouver General Hospital. The new OGM platform iteration, Stratys, incorporates updates aimed at increasing throughput and is under evaluation to facilitate ramp-up of OGM-based testing in our clinical laboratory. We describe preliminary results of a clinical verification of Stratys, through comparison with the Saphyr, in diagnostic evaluation of hematologic malignancies.</div></div><div><h3>Methods</h3><div>We processed and annotated 10 specimens from patients with AML (6), ALL (3), and MDS (1) using manufacturer-recommended protocols on Saphyr (rare variant analysis pipeline) and Stratys (low allele fraction guided assembly pipeline). Data quality metrics and identified variants (filtered using provided control datasets and laboratory-developed gene-based filtering approaches) were compared between platforms.</div></div><div><h3>Results</h3><div>All 32 variants meeting our laboratory's clinical reporting standards using Saphyr were detected using Stratys. Data quality metrics differed negligibly for cases between platforms. Unfiltered Stratys data included substantially more duplication calls (median 64% increase over Saphyr) but population-filtered variants were similar in number between platforms. In representative variants (including translocations/deletions/chromosome gain/loss), breakpoint nucleotide positions were typically identical or, infrequently, differed by less than 10kb interplatform, and variant frequency measures typically differed by 8% on average (without a clear interplatform bias).</div></div><div><h3>Conclusion</h3><div>Preliminary analysis of verification data indicates Stratys to perform comparably to Saphyr for detection of reportable somatic variants in clinical bone marrow specimens, albeit with relative changes in variant call rates (for variants we currently consider non-reportable) and reported variant burdens.</div></div>","PeriodicalId":49225,"journal":{"name":"Cancer Genetics","volume":"286 ","pages":"Page S2"},"PeriodicalIF":1.4,"publicationDate":"2024-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142322864","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-01DOI: 10.1016/j.cancergen.2024.08.007
Wenhua Zhou, Eric Fredrickson, Maria Longhurst, Emily Aston, Bo Hong
The AneuVysion FISH probe kit is FDA approved for detecting aneuploidy of chromosomes X, Y, 13, 18 and 21 on amniocytes (AF). Clinically, this kit is also used to identify numerical abnormalities in other specimen types, including chorionic villi sampling (CVS), cord blood, and peripheral blood. Currently, the majority of FISH tests in our lab have been adapted to the automatic slide processing system - BioDot instruments. Therefore, a validation was conducted to integrate the procedure of aforementioned Aneuploidy FISH test into the BioDot instruments. During the development phase, 34 cases were processed following the standard BioDot protocol with the exception of AF and CVS. Cell pellets from AF and CVS being manually applied onto slides due to the low cellularity. The FISH signal intensity and background were compared based on a 0 -7 scale. The FISH signal intensity was comparable between the slides processed by Biodot protocol and the lab current protocol (5.0 vs 4.8, p =4.8); the background score improved on the slides with the Biodot procedure (5.3 vs 5.1, p = 0.01). Subsequently, 21 cases for accuracy and 12 cases for precision (between run and within run) were tested on BioDot instruments for validation. These cases demonstrated 100% concordant FISH results with those from the current manual procedures. This study suggests that the traditional FDA-approved FISH Aneuvysion test can be seamlessly adapted to our standard Biodot FISH slide processing. This transition will streamline lab workflow, increase work efficiency, better FISH test quality and improve cost effectiveness.
AneuVysion FISH 探针试剂盒经 FDA 批准用于检测羊膜细胞（AF）上 X、Y、13、18 和 21 染色体的非整倍体。在临床上，该试剂盒还可用于鉴定其他类型标本中的染色体数目异常，包括绒毛取样（CVS）、脐带血和外周血。目前，我们实验室的大多数 FISH 检测都已适用于自动玻片处理系统 - BioDot 仪器。因此，我们进行了一项验证，将上述非整倍体 FISH 检测程序整合到 BioDot 仪器中。在开发阶段，除 AF 和 CVS 外，按照标准 BioDot 方案处理了 34 个病例。由于 AF 和 CVS 中的细胞颗粒较少，因此采用手工将细胞颗粒贴在载玻片上。FISH 信号强度和背景根据 0-7 级进行比较。采用 Biodot 方案和实验室现行方案处理的玻片的 FISH 信号强度相当（5.0 vs 4.8，p =4.8）；采用 Biodot 方案处理的玻片的背景得分有所提高（5.3 vs 5.1，p =0.01）。随后，在 BioDot 仪器上对 21 个病例的准确性和 12 个病例的精确性（运行间和运行内）进行了验证测试。这些病例的 FISH 结果与当前手动程序的结果 100%一致。这项研究表明，传统的 FDA 批准的 FISH Aneuvysion 检验可以无缝地适应我们的标准 Biodot FISH 玻片处理。这一转变将简化实验室工作流程、提高工作效率、改善 FISH 检测质量并提高成本效益。
{"title":"5. Implementation of Automatic Slide Processing for Aneuploidy FISH Test","authors":"Wenhua Zhou, Eric Fredrickson, Maria Longhurst, Emily Aston, Bo Hong","doi":"10.1016/j.cancergen.2024.08.007","DOIUrl":"10.1016/j.cancergen.2024.08.007","url":null,"abstract":"<div><div>The AneuVysion FISH probe kit is FDA approved for detecting aneuploidy of chromosomes X, Y, 13, 18 and 21 on amniocytes (AF). Clinically, this kit is also used to identify numerical abnormalities in other specimen types, including chorionic villi sampling (CVS), cord blood, and peripheral blood. Currently, the majority of FISH tests in our lab have been adapted to the automatic slide processing system - BioDot instruments. Therefore, a validation was conducted to integrate the procedure of aforementioned Aneuploidy FISH test into the BioDot instruments. During the development phase, 34 cases were processed following the standard BioDot protocol with the exception of AF and CVS. Cell pellets from AF and CVS being manually applied onto slides due to the low cellularity. The FISH signal intensity and background were compared based on a 0 -7 scale. The FISH signal intensity was comparable between the slides processed by Biodot protocol and the lab current protocol (5.0 vs 4.8, p =4.8); the background score improved on the slides with the Biodot procedure (5.3 vs 5.1, p = 0.01). Subsequently, 21 cases for accuracy and 12 cases for precision (between run and within run) were tested on BioDot instruments for validation. These cases demonstrated 100% concordant FISH results with those from the current manual procedures. This study suggests that the traditional FDA-approved FISH Aneuvysion test can be seamlessly adapted to our standard Biodot FISH slide processing. This transition will streamline lab workflow, increase work efficiency, better FISH test quality and improve cost effectiveness.</div></div>","PeriodicalId":49225,"journal":{"name":"Cancer Genetics","volume":"286 ","pages":"Page S2"},"PeriodicalIF":1.4,"publicationDate":"2024-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142322866","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-01DOI: 10.1016/j.cancergen.2024.08.063
Binish Narang
Despite significant advances in cancer research, cancer remains a major public health concern, with breast cancer being one of the leading causes of death among women. The mitogen-activated protein kinase kinase kinase 1 (MAP3K1) codes for a serine/threonine kinase abundant in the c-Jun N-terminal kinase, mitogen-activated protein kinase, and Nf-kappa-β pathways, which are involved in tumorigenesis. Multiple statistical tests were conducted on the TCGA and METABRIC datasets downloaded from cBioPortal to analyze MAP3K1's relevance in breast cancer. Other tools, including TIMER 2.0, Kaplan-Meier Plotter, UALCAN, and STRING, were implemented to provide additional insight into MAP3K1 in different types of omics data. Results revealed that, though MAP3K1 alterations are relatively uncommon overall, they are most common in breast cancer. These alterations mostly included truncating mutations and often co-occurred with alterations in PIK3CA, an already established biomarker in breast cancer research. Survival analysis indicated that MAP3K1 underexpression was strongly associated with lower patient survival. MAP3K1 was underexpressed for African Americans, triple-negative breast cancer patients, and stage 4 patients, while its phosphoprotein was overexpressed for these demographics. Drug targets or other targeted therapy options that limit MAP3K1 phosphoprotein expression could potentially improve patient outcomes, especially for the aforementioned demographics. However, limited information is known about this phosphoprotein, so there is an unmet need to address this lack of knowledge and eventually find ways to combat its excessive expression in breast cancer.
{"title":"61. MAP3K1 identified as a prognostic biomarker in breast cancer after multi-omics bioinformatics analysis","authors":"Binish Narang","doi":"10.1016/j.cancergen.2024.08.063","DOIUrl":"10.1016/j.cancergen.2024.08.063","url":null,"abstract":"<div><div>Despite significant advances in cancer research, cancer remains a major public health concern, with breast cancer being one of the leading causes of death among women. The mitogen-activated protein kinase kinase kinase 1 (<em>MAP3K1</em>) codes for a serine/threonine kinase abundant in the c-Jun N-terminal kinase, mitogen-activated protein kinase, and Nf-kappa-β pathways, which are involved in tumorigenesis. Multiple statistical tests were conducted on the TCGA and METABRIC datasets downloaded from cBioPortal to analyze <em>MAP3K1</em>'s relevance in breast cancer. Other tools, including TIMER 2.0, Kaplan-Meier Plotter, UALCAN, and STRING, were implemented to provide additional insight into <em>MAP3K1</em> in different types of omics data. Results revealed that, though <em>MAP3K1</em> alterations are relatively uncommon overall, they are most common in breast cancer. These alterations mostly included truncating mutations and often co-occurred with alterations in PIK3CA, an already established biomarker in breast cancer research. Survival analysis indicated that <em>MAP3K1</em> underexpression was strongly associated with lower patient survival. <em>MAP3K1</em> was underexpressed for African Americans, triple-negative breast cancer patients, and stage 4 patients, while its phosphoprotein was overexpressed for these demographics. Drug targets or other targeted therapy options that limit <em>MAP3K1</em> phosphoprotein expression could potentially improve patient outcomes, especially for the aforementioned demographics. However, limited information is known about this phosphoprotein, so there is an unmet need to address this lack of knowledge and eventually find ways to combat its excessive expression in breast cancer.</div></div>","PeriodicalId":49225,"journal":{"name":"Cancer Genetics","volume":"286 ","pages":"Page S20"},"PeriodicalIF":1.4,"publicationDate":"2024-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142323476","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号