Resource Library

Showing 1 - 5 of 79 results.

AGBTPH19 - Application of Lean Manufacturing Methodologies in High Throughput Genomic Sequencing

Peter Trefry1, Sam DeLuca1, Mike Nasuti1, Shannon Adams1, Marissa Gildea1, Doug Gobron1, Tom Howd1, Tim DeSmet1


1Broad Institute of MIT and Harvard, Cambridge, MA


The utility and application of genomics to understand disease, and the continuing trend to utilize genomics in healthcare, results in an ever increasing demand for greater sequence data generation. Despite the significant reductions in per-base sequencing cost over the last decade, the infrastructure, capital, and reagent costs are still relatively expensive. Top of the line sequencers can cost over 1 million dollars per instrument, and sequencing run costs can still be tens of thousands of dollars. With such high fixed cost associated with genome data generation, it is important to maximize capacity utilization and reduce the non-value add and wasteful workflow process steps. We demonstrate the application of lean manufacturing methodologies and visual management techniques to the genomic sequencing workflow, which results in achieving a sequencer utilization rate of around 90%, while three fold scaling our library preparation process to over 300,000 samples destined for exome and whole human genome sequencing annually.


By combining the sample preparation methods for both exome and whole genome sequencing into a unified, modularized workflow, samples and reagent supply chains can be optimized resulting in more efficient, and cost effective processing. Additional benefits include reductions to work in process and overall cycle times. Here, we illustrate the methodologies that enable low cost per base sequence data generation applicable across large sequencing cores, and modest sized data generation groups.

AGBTPH19 - Clinical Validation of Illumina Array-based Genotyping for the All of Us Research Program

Michael DaSilva1, Alyssa Macbeth1, Steven Harrison1, Betty Wolff1, Maegan Harden1, Gina Vicente1, Sarah Babchuck1, Katelyn Flowers1, Kunsang Gyaltsen1, Brian Granger1, George Grant1, Heidi Rehm1, Stacey Gabriel1, Niall Lennon1


1Broad Institute of MIT and Harvard, Cambridge, MA


As the utility of genomics evolves, a need has emerged for large-scale genomic datasets produced in a clinical setting. To satisfy that niche, the Genomics Platform at the Broad Institute has established CAP/CLIA compliant genotyping capabilities at scale. Once such effort requiring these capabilities is the All of Us research project (AoURP). Funded by the National Institutes of Health (NIH), AoURP aims to generate sequencing and genotyping data from 1 million or more research participants across the U.S. Medically actionable results from a pre-defined set of genes (termed the AoU Medically Actionable Panel, AoUMAP) will be returned to participants after orthogonal validation in a clinical validation laboratory. Genotyping will be run using the Illumina All of Us (AoU) array that has been specifically designed with 1.8 million markers across the human genome. AoU array content has been designed to capture pathogenic and likely pathogenic sites (as defined by ClinVar) across the AoUMAP, in addition to pharmacogenetic markers.

As one of three Genome Centers selected by the AoURP program, the Broad Institute Genomics Platform has built the operational and clinical infrastructure required to genotype >300,000 AoURP participants over the next 4-5 years. We performed an analytical validation study to assess the accuracy and precision of the AoU Array. A combination of reference samples, PGx cell lines, and previously tested clinical samples were used to verify that the array is suitable for its intended use as part of the AoURP. In addition, we have established the capacity to meet AoURP scale, while running other concurrent, large-scale projects.

Here we discuss the onboarding process for the new array, the establishment of CLIA/CAP quality standards in our genotyping process, and the results of our analytical validation study.

AGBTPH19 - Navigating the Regulatory Landscape for All of Us Genotyping and Whole Genome Sequencing Processes

Niall Lennon1, Kim Doheny2, Donna Muzny3, Christina Lockwood4, Ginger Metcalf3 on behalf of the members of the All of Us Regulatory Working Group


1Broad Institute of MIT and Harvard, Cambridge, MA; 2Johns Hopkins University, Baltimore, MD; 3Baylor College of Medicine, Houston TX; 4Northwest Genomics Center, University of Washington, WA


The All of Us Research Program (AoURP) is a large collaborative initiative sponsored by the National Institutes of Health (NIH) with a primary objective of building a research resource composed of participant-provided information (PPI), including environmental, physiologic, and health data and biospecimens from 1 million or more research participants who reflect the diversity of the U.S. Participants are also invited to undergo physical measurements and provide biospecimens from which genomic information and other biomarkers will be derived. 


A core value of the program is that participants will have access to their data and that they may receive information potentially relevant to their own health. To this end, the program will include a return of results arm, in which predispositions to the development of certain diseases will be assessed through examination of a panel of genes (termed the AoU Medically Actionable Panel) and high confidence pharmacogenetic variants. Primary testing will involve Whole Genome Sequencing (WGS) as well as a custom genotyping array (AoU Array). Both assays have been validated as lab-developed tests (LDTs) by a group of CAP/CLIA certified clinical labs in genome centers across the US (Broad Institute, Baylor College of Medicine, Johns Hopkins, and the UW Northwest Genome Center). 


Since health-related and potentially actionable information is being returned to healthy individuals who have not had a physician-ordered clinical test for a specific condition, the US FDA has determined that this represents a high risk activity and therefore requires an Investigational Device Exemption (IDE). We present here the process of navigating the IDE pre-submission and submission process for this large, multi-center genomic testing project.

AGBTPH19 - Comparison of Blood Collection Vacutainers for Reduction of gDNA Contamination in cfDNA Studies

Emily Moore1, Evan McDaid1, Katie Larkin1, Michelle Cipicchio1, Nicholas Fitzgerald1, Michael Nasuti1, Brendan Blumenstiel1, Tim DeSmet1, Viktor Adalsteinsson1, Niall Lennon1, Stacy Gabriel1 


1Broad Institute of MIT and Harvard, Cambridge, MA


Liquid biopsy and cfDNA sequencing allows for rapid analysis of cancer progression in the field of precision medicine. Successful library construction of a cfDNA sample is dependent upon the ability to purify cfDNA in the absence of gDNA. Throughout the sample collection and handling process, liquid biopsy samples may be subjected to cell lysis, resulting in the release of gDNA, negatively affecting sample integrity. Therefore, controlling variables in the sample collection process is an important factor in improving the resulting variant detection. Vacutainers and the various stabilizers within them function to postpone gDNA contamination. As minimizing gDNA presence is critical for successful sequencing, it is important to utilize a vacutainer that provides the greatest degree of sample integrity. 


We evaluated seven vacutainers across multiple time points over a 32-day period to determine their relative stability. Whole blood collected from four unique healthy donors was fractionated at predetermined time intervals. The comparison was performed over a large window of time which allowed for changes in the relative abundance of gDNA to be identified, as well as determining success in ultra-low-pass whole-genome sequencing at 0.1X coverage. 


Analysis of concentration, DNA fragment size, and sequencing results were used to determine vacutainer performance in reducing gDNA contamination. Through this analysis we observed there is a relationship between vacutainer type and degree of gDNA contamination over time. Collecting this information will allow for a recommendation on which tube type should be utilized for blood collection, as well as the optimal time frame from collection to fractionation to avoid gDNA contamination, improving a sample’s chance of success in sequencing. Fine-tuning this aspect of sample collection will be particularly impactful for clinical applications where sample integrity and mutation detection will ultimately have an effect on a patient’s diagnosis, treatment and disease outcome.

AGBTPH19 - A Scalable Liquid Biopsy Pipeline Using Duplex Sequencing

Mark Fleharty1, Madeleine Duran1, Matt DeFelice1, Brenden Blumensteil1, Carrie Cibulskis1, Viktor Adalsteinsson1, Stacey Gabriel1, Niall Lennon1


1Broad Institute of MIT and Harvard, Cambridge, MA


Liquid biopsies are a relatively non-invasive procedure enabling greater access to patient specimens and allows for the characterization of somatic variants repeatedly over time. With this ability to accurately detect somatic events, liquid biopsy has the potential to significantly impact the course of precision medicine in cancer.


We present an end-to-end pipeline that delivers high quality liquid biopsy results to support translational research.  Using a custom data pipeline and a lab process that incorporates duplex unique molecular indices (UMI) we have benchmarked a 396 gene pan-cancer panel, a multiple myeloma panel, and glioma panel. This pipeline makes use of UMIs for increasing the available depth of reads and reduces error by utilizing duplex-consensus called reads.  We have benchmarked this technology using pooled sample analysis to simulate somatic variants from a tumor and normal-normal analysis as an independent measure of false positive rate. Our pipeline produces duplex consensus called bams and variant calls. We have developed a set of novel variant filters specific to duplex consensus calls that are not found in other types of somatic pipelines.  With these novel filters, our liquid biopsy pipeline achieves a false positive rate under 0.5 per megabase with > 90% sensitivity at 1% allele fraction.