Next generation sequencing (NGS) has revolutionized genomics and is now being widely adopted for clinical sequencing. Whole exome or whole genome sequencing (WES/WGS) is the ultimate genetic test and many success stories provide a taste of its power. Targeted NGS-based gene panels are typically an order of magnitude smaller than WES/WGS-based testing but follow the same principles. However, while the cost of generating high-quality whole genome sequence data is rapidly dropping, analysis of the enormous number of variants detected is still very complex, and a task of annotating NGS data for clinical grade reporting and interpretations remain a challenge. Cambridge Healthtech Institute’s Fourth Annual NGS Knowledgebases, Annotation, and Interpretation symposium is designed to discuss best practices in NGS data analysis, and NGS results annotation, interpretation, disclosure and applications.
Thursday, February 23
7:00 am Registration and Morning Coffee
8:25 Chairperson’s Opening Remarks
Mark E. Nunes, M.D., Associate Professor, Pediatrics, Division Chief, Medical Genetics, Kaiser Permanente
8:30 Data Sharing through the NCI Genomics Data Commons
Louis M. Staudt, M.D., Ph.D., Director, Center for Cancer Genomics, Co-Chief, Lymphoid Malignancies Branch, National Cancer Institute, National Institutes of Health
The National Cancer Institute (NCI) Genomic Data Commons (GDC) has been established to promote the sharing of cancer genomic and clinical data, with the aim to foster precision medicine approaches to the diagnosis and treatment of cancer. The GDC is being developed in partnership with investigators the University of Chicago and the Ontario Institute for Cancer Research and currently contains data from over 14,000 patients with cancer. The GDC not only houses data collected from NCI-sponsored programs but is also open for any investigator who wishes to share cancer genomic and clinical data broadly. As the number of cases in the GDC grows, its explanatory power will increase, enabling the identification of low-frequency genetic drivers of cancer, an understanding of genomic determinants of therapeutic response, and potentially the constitution of clinical trial cohorts of patients with shared genetic lesions.
9:00 Has the Genomic Infrastructure Been Built to Allow Precision Medicine?
Mark E. Nunes, M.D., Associate Professor, Pediatrics, Division Chief, Medical Genetics, Kaiser Permanente
The Human Genome Project converged with the Digital Age to create the large-scale infrastructure needed to generate and analyze genomic “big data”. The community infrastructure for sharing genomic data, and the personal/personnel infrastructure to exploit genomic data, have lagged. A clinician’s perspective on the state of community and personal infrastructure, as providers and patients interface with electronic medical records, will ask: enough to overcome the personnel and knowledge gaps?
9:30 Free the Data: One Laboratory’s Approach to Knowledge-Based Genomic Variant Classification
Madhuri Hegde, Ph.D., FACMG, Adjunct Professor, Emory University, Vice President and Chief Scientific Officer, Global Laboratory Services, Diagnostics, PerkinElmer, Inc.
High quality accurate classification of the clinical significance of sequence variant identified is essential in releasing the full potential of genomic medicine. The amount of sequence generated within clinical laboratories has increased dramatically with the advent of lower-cost, more automated Sanger sequencing and next-generation sequencing technologies. We have developed a variant curation interface management suite: EmVar and EmVClass, which is used to store variants and facilitate variant classification.
10:00 Systematic Assessment of Clinical Actionability Associated with Genomic Variation
Elizabeth Webber, MS, Research Associate, Center for Health Research, Kaiser Permanente
ClinGen’s Actionability Working Group developed a structured framework to assess clinical actionability of genes and associated disorders based upon the availability of clinical interventions that could improve future health outcomes in patients and their at-risk relatives. This framework provides support to the research and clinical community for making clear, streamlined, and consistent determinations of clinical actionability based upon transparent criteria to guide analysis and reporting of variation in genome-scale sequencing.
10:30 Coffee Break with Exhibit and Poster Viewing
11:10 Chairperson’s Remarks
Birgit H. Funke, Ph.D., FACMG, Associate Professor, Pathology, MGH/Harvard Medical School; Director, Clinical Research and Development, Laboratory for Molecular Medicine - Partners HealthCare
11:15 Developing Knowledge Resources for the Diagnostic Lab Director
Birgit H. Funke, Ph.D., FACMG, Associate Professor, Pathology, MGH/Harvard
Medical School; Director, Clinical Research and Development, Laboratory for Molecular Medicine - Partners HealthCare
Exome and genome sequencing will eventually serve as a first line test for many genetic disorders. While the technology is no longer a barrier, developing and updating knowledge resources on large numbers of genes, variants and disorders is a major bottleneck. This presentation will discuss national efforts to develop and curate knowledge resources that will guide diagnostic laboratory directors in all aspects of genomic sequencing including test design, validation and interpretation.
11:45 SOAR: Scalable OMICS Analysis and Reporting
Andrew Stubbs, Ph.D., Assistant Professor, Bioinformatics, Erasmus University Medical School
We have implemented, SOAR, a generalized and scalable open source solution for integrated analysis of clinical and "OMICs" data. SOAR provides biomarker discovery and validation, with the latest "OMICs" tools available via Galaxy (use.galaxy.org) to the medical researchers in clinical and translational research projects. SOAR uses our GalaxyFlow to access multiple Galaxy instances (including tools) via a single graphical user interface. GalaxyFlow ensures that SOAR platform provides FAIR (Findable, Accessible, Interoperable, and Reusable) data principles where possible. The utility of SOAR will be demonstrated with existing translational research projects at the Erasmus University Medical Center.
12:15 pm Enjoy Lunch on Your Own
1:50 Chairperson’s Remarks
Susan Mockus, Ph.D., Manager, Clinical Analytics & Curation, The Jackson Laboratory for Genomic Medicine
2:00 Using Exome Aggregation (ExAC) Dataset for the Interpretation of Rare Variants in Mendelian Diseases
Monkol Lek, Ph.D., Research Fellow, Massachusetts General Hospital
Large-scale reference data sets of human genetic variation are critical for the medical and functional interpretation of DNA sequence changes. The ExAC data set contains variants from over 120,000 individuals aggregated from a variety of large-scale sequencing projects. In this presentation, we will provide a general overview of the production of the ExAC data set, recent updates and also discuss examples of how the data set has been used for the interpretation of rare variants and development of methods available to the community.
2:30 The 100,000 Genomes Project: Transforming the UK’s National Health Service
Joanne Mason,Ph.D., Director of Sequencing and Sample Acquisition, Genomics England, Queen Mary University of London
The 100,000 genomes project is transforming the UK’s National Health Services introducing whole genome sequencing as a standard of care test for rare disease and cancer patients. My talk will cover our approach and infrastructure to deliver this transformational program for patients in England and approaches to interpreting whole genome data on a population scale.
3:00 Tackling the Task of Precision Immuno-Oncology in an Integrated Health System
Terence Rhodes, M.D.,Ph.D., Director of Immuno-Oncology, Intermountain Healthcare
Although an active area of research, the lack of current clinical biomarkers for cancer immunotherapy comes at a significant delay in effective treatments and increased costs for the majority of patients who do not benefit from immunotherapy. Intermountain Healthcare’s unique resources will play a role in personalizing immuno-oncology.
3:30 Refreshment Break and Poster Competition Winner Announced in the Exhibit Hall
4:15 Creating Standards and Transparency for Interpretation of Somatic Variants
Susan Mockus, Ph.D., Manager, Clinical Analytics & Curation, The Jackson Laboratory for Genomic Medicine
Genomic tumor profiling enables insights into prognostic, diagnostic, and predictive biomarkers of disease. Somatic variants are often characterized by clinical actionability. Actionability is defined in a variety of ways ranging from variant pathogenicity to connection to targeted therapies and available clinical trials. Tier ranking systems have been commonly implemented, but none uniformly adopted. The need for interpretation standards and methods for providing transparency will be described.
4:45 Pushing the Limits - Challenges of Somatic Variant Detection
Robert Daber, Ph.D., Founder and CEO, Gnosity Consults
NGS continues to emerge as a powerful diagnostic methodology for the characterization of mutation status in a variety of tumors. Here we describe our experience detecting various complicated somatic mutation types across a variety of tumor types, including both low allele frequency mutations as well as complicated insertion and deletion events. We will also discuss our strategies for decreasing the lower limit of detection by customizing the limits of detection for each genomic loci independently.
5:15 Large-Scale, Cloud-Based Analysis of Cancer Genomes: Lessons Learned from the PCAWG Project
Brian O’Connor, Technical Director, Analysis Core Genomics Institute, UCSC
The PanCancer Analysis of Whole Genomes (PCAWG) project is a large-scale, highly distributed research collaboration designed to identify common patterns of mutations across 2,800 cancer genomes. The use of clouds, both public and private, was instrumental in analyzing this dataset using current best practice pipelines. This talk describes the technical infrastructure built for the project, how we leveraged cloud environments to perform the “core” analysis, and the lessons learned along the way. It will also explore the nature of the dataset and how it can be leveraged to support research and clinical applications.
6:45 Close of Day
Friday, February 24
7:00 am Breakfast Presentation (Sponsorship Opportunity Available) or Morning Coffee
8:00 Registration Open
8:25 Chairperson’s Remarks
Catherine Brownstein, Instructor, Genetics and Genomics, Boston Children’s Hospital and Harvard Medical School
8:30 Annotation and Interpretation of Clinical Exome Sequencing
Wayne W. Grody, M.D., Ph.D., Professor, Medical Genetics and Molecular Pathology, Pathology & Lab Medicine, Pediatrics, Human Genetics, Director, Molecular Diagnostic Laboratories and Clinical Genomics Center, University of California Los Angeles School of Medicine
Our center has been performing clinical-grade whole-exome sequencing (WES) for the diagnosis of rare Mendelian disorders since January 2012. In addition to our in-house bioinformatics pipeline and externally available databases and algorithms, all mutations and variants are interpreted by a unique “Clinical Genomics Board” comprised of lab directors, technologists, bioinformaticists, genetic counselors, medical geneticists, and the ordering clinicians. We find that this approach provides the most “value-added” clinical insight for proper annotation and reporting of variants.
9:00 CLARITY Undiagnosed – Interpreting Clinical Variation
Catherine Brownstein, Instructor, Genetics and Genomics, Boston Children’s Hospital and Harvard Medical School
Variant interpretation is a rapidly evolving field; there remains a great diversity in application of criteria and choices of genes and variants to be reported when clear cut pathogenic mutations are not immediately obvious. The CLARITY Challenges are an effective means for assessing current practices for using next-generation sequencing. The data collected during this contest will be used to accelerate broad dissemination of diagnostic sequencing practices that are suitable for clinical use.
9:30 Building a Framework for Consistent and Accurate Clinical Interpretation of Germline Sequence Variants
Keith Nykamp, Ph.D., Senior Scientist, Clinical Genomics Group, Invitae
With the ACMG ISV guidelines as a starting point, we developed a weighted, score-based classification system designed to be scalable across a large team of variant scientists. This system, which we call Sherloc, was implemented in our clinical reporting workflow and iteratively revised based on our experience with more than 15,000 variants. This presentation will discuss some of the challenges we encountered, and how the Sherloc system overcomes these challenges.
10:00 Mendel, meet Mendeleev: Why Genotypes Matter More than Variants -- and What You Can Do About It
Nathaniel Pearson, Ph.D., Senior Director, Scientific Engagement & Public Outreach, New York Genome Center
To be Announced
10:30 Coffee Break with Exhibit and Poster Viewing
11:15 Adaptive Genomics Engine (AGE) for Cloud-Based Machine Learning of Cell-Free DNA (cfDNA) to Enable Computational Classification and Biomarker Discovery in Cancer Research
Gabriel Otte, Ph.D., CEO, Co-Founder, Freenome, Inc.
The underlying biology of cell-free DNA (cfDNA) and projected signatures in circulation are fundamentally different from tissue. Thus, tools are needed to facilitate the deconvolution of these novel cfDNA signatures. Freenome built the Adaptive Genomics Engine (AGE) for read-level transformation of cfDNA sequence data to enable computational classification with deep learning. AGE generates unique feature representations of cfDNA to define data structures independent of traditional mutation calling such as read length distributions. This method showed >95% accuracy for disease detection (n = 351) and estimated a reduced dimensional subspace that preserves divergence between tissues of origin in both early- and late-stage lung and prostate cancers, indicative of multi-cancer discrimination with AGE.
11:45 The Human Microbiome: Data Challenges and Solutions
Andreas M. Kogelnik, M.D., Ph.D., Open Medicine Institute
Examination of various human microbiomes is yielding valuable, clinically-relevant information; however, there is still much to learn. Human microbiome analysis is the study of microbial communities found in and on the human body. The goal of human microbiome studies is to understand the role of microbes in health and disease. High throughput methods have enabled increasingly relevant studies with increasing clinical impact that is both surprising and broad-reaching at times. There remains enormous work to be done for data analysis and for application of these technologies.
12:15 pm Public and Private Databases: Competition or Cooperation
Moderator: Catherine Brownstein, Instructor, Genetics and Genomics, Boston Children’s Hospital and Harvard Medical School
Panelists: Irene C. Blat, Ph.D., Scientific Director of Translational Genomics, Application Sciences, WuXi NextCODE Genomics
Nathaniel Pearson, Ph.D., Senior Director, Scientific Engagement & Public Outreach, New York Genome Center
Wayne W. Grody, M.D., Ph.D., Professor, Medical Genetics and Molecular Pathology, Pathology & Lab Medicine, Pediatrics, Human Genetics, Director, Molecular Diagnostic Laboratories and Clinical Genomics Center, University of California Los Angeles School of Medicine
- Is there a single nomenclature source for genes and variants that should be adopted across all databases?
- How should databases provide transparency to data collection sources and evaluation processes?
- What levels of evidence should be required for a new variant?
12:45 Close of Symposium