Best practices for the analytical validation of clinical whole-genome sequencing intended for the diagnosis of germline disease

Abstract

Whole-genome sequencing (WGS) has shown promise in becoming a first-tier diagnostic test for patients with rare genetic disorders; however, standards addressing the definition and deployment practice of a best-in-class test are lacking. To address these gaps, the Medical Genome Initiative, a consortium of leading healthcare and research organizations in the US and Canada, was formed to expand access to high-quality clinical WGS by publishing best practices. Here, we present consensus recommendations on clinical WGS analytical validation for the diagnosis of individuals with suspected germline disease with a focus on test development, upfront considerations for test design, test validation practices, and metrics to monitor test performance. This work also provides insight into the current state of WGS testing at each member institution, including the utilization of reference and other standards across sites. Importantly, members of this initiative strongly believe that clinical WGS is an appropriate first-tier test for patients with rare genetic disorders, and at minimum is ready to replace chromosomal microarray analysis and whole-exome sequencing. The recommendations presented here should reduce the burden on laboratories introducing WGS into clinical practice, and support safe and effective WGS testing for diagnosis of germline disease.

Rare skeletal disorders: a multidisciplinary postnatal approach to diagnosis and management

“…In addition, chromosomal microarray (CMA) is applied to exclude unbalanced chromosomal aberrations. In the near future, novel techniques, such as whole genome sequencing (WGS), will presumable replace CMA, CES, PA and WES due to diagnostic superiority [ 2 , 14 ].…”

“…In recent years, these novel genetic techniques have become the first-tier diagnostics for rare monogenetic phenotypes and therefore have replaced a detailed clinical investigation and more invasive diagnostic procedures in many medical disciplines [ 14 ]. In contrast interpretation of genetic data in skeletal disorders often requires detailed clinical and radiological phenotyping, implying a combined clinical, radiological and genetic workup as the best first-tier diagnostic approach in skeletal disorders [ 2 , 12 ].…”

Clinical utility of genomic sequencing: a measurement toolkit

“… Fryback & Thornbury domain & definition Measurement construct Indicator Data source Data collection strategy (and examples) References Level 1: Technical efficacy In the laboratory, does the test measure what it purports to measure? Analytic validity Sensitivity (recall), specificity, precision (technical positive predictive value) when “gold standard” reference available Positive percent agreement (PPA), negative percent agreement (NPA) when only standard technical results available Laboratory Information System or Lab report Marshall 27 FDA 81 Level 2: Diagnostic accuracy efficacy Does the test result distinguish patients with and without the target disorder? Clinical validity Variant classification accuracy Genotype–phenotype matching Test outcome (positive, negative, inconclusive) Laboratory Information System or Lab report Rehm 38 Strande 82 Richards 83 Level 3: Diagnostic thinking efficacy Does the test help a clinician to come to a diagnosis?…”

“…We retain the levels of technical and diagnostic accuracy efficacy (i.e., levels 1 and 2) as essential starting points in our guiding framework as they are fundamental precursors to achieving clinical utility. However, since these laboratory-based components of efficacy are well-debated and described in the WGS literature and in recent guidelines published by members of our group 27 , we focus here on four levels of the efficacy model (i.e., levels 3–6) that align most directly with a broad definition of clinical utility and extend beyond laboratory-based components of efficacy. In emphasizing these four levels of efficacy as components of clinical utility, our intent is to encourage the use of a broad set of health and non-health-related indicators of value to bolster the state of evidence in this area, rather than to convey that all aspects of clinical utility need to be achieved for WGS adoption and reimbursement.…”

Dysgu: efficient structural variant calling using short or long reads

“…Analysis of structural variants (SVs) with whole genome or targeted enrichment sequencing is used in the clinic for diagnosing acquired or inherited genetic diseases (1) and for investigating mechanisms of genomic complexity in cancer and other pathologies (2–6). Sequencing using short paired-end reads (PE) is well established for genomic analysis due to mature workflows and low sequencing costs, although increasingly, long-read (LR) sequencing technologies are being utilized for these purposes.…”