Life Sciences

Whitepaper: Improving the diagnosis of infectious diseases with microbial genomic analysis

Issue link: https://read.uberflip.com/i/1331299

Contents of this Issue

Navigation

Page 1 of 2

2 a comprehensive view of the microbial system," he said. "As long as you employ an agnostic sequencing method, you get to explore the wide range of microorganisms living in the body." The microbial cell-free DNA represents microbes from all body systems, not just the bloodstream, enabling the test to detect infections regardless of where they might reside. Additionally, the test is quantitative, so the sicker the individual, the more microbial DNA will be present, and as a treatment takes effect, the abundance of microbial cell-free DNA will decrease. The comprehensive snapshot provided by the .DULXV 7HVW VLJQLƓFDQWO\ VKRUWHQV WKH WLPH WR diagnosis, with typical turnaround times of about a day. This differs from the status quo, in which serial testing of patients leads to "a prolonged diagnostic odyssey," Dr. Bercovici said. "At the end of the day, it doesn't come down to how fast a single test runs, but rather, how fast the physician can put the patient on the right course of treatment." Clinical utility for the Karius Test has been demonstrated in multiple peer-reviewed publications, including an analytical and clinical validation landmark study published in Nature Microbiology 1 . "We had clinical validity assessment across multiple patient types," said Dr. Bercovici, including pediatric patients and those with compromised immune systems, who particularly EHQHƓWIURPIDVWHUWDUJHWHGWUHDWPHQW Because the sequencing-based microbial cell- free DNA test is comprehensive, analyzing the data requires much more computing power than previous tests. "Conventional infectious disease diagnostic methods are only scratching the surface of what exists in our deep data," said Dr. Bercovici. As the molecular technology advances, it must be accompanied by strong data technology to create value. For its platform, Karius combines a continuously expanding genomic database with machine- learning algorithms. The database uses a "human- in-the-loop" methodology, in which human experts review genomic datasets as curated by algorithms. On the machine-learning side, the platform applies a series of algorithms aimed at identifying the VSHFLƓFSDWKRJHQVGULYLQJWKHGLVHDVHIURPWKH raw sequencing data. By leveraging Amazon Web Services (AWS), Dr. Bercovici said, Karius has been able to focus on developing its lab and analysis technologies, rather than devoting resources to IT. "AWS as our partner provided us with all the tools we need in order to reduce the engineering burden surrounding dealing with such a complex architecture," he said. Furthermore, AWS provides the high level of data security necessary for dealing with sensitive PHGLFDOLQIRUPDWLRQŏ,ōPFRQƓGHQWWKHVHFORXG services offer a more than adequate solution to any security concerns when it comes to providing a HIPAA compliant service," he said. Dr. Bercovici walked through the details of the Karius Test. First, the DNA sequence data are cleaned up and compared against a human reference genome. Sequences that map to an H[WHQGHG KXPDQ UHIHUHQFH DUH ƓOWHUHG RXW A GENOMEWEB / AWS WHITE PAPER KARIUS TEST PROCESS Reports are typically available one day after specimen receipt JANUARY 2020

Articles in this issue

view archives of Life Sciences - Whitepaper: Improving the diagnosis of infectious diseases with microbial genomic analysis