Development of a methods-based proficiency test for next generation sequencing (NGS). N. Aziz1, J. Durtschi2, Q. Zhao1, L. Bry12, D. Driscoll1, J. Gibson15, W. Grody14, M. Hedge10, G. Hoeltge16, 17, 18, D. Leonard13, J. Merker6, L. Palicki1, R. S. Robetorye7, I. Schrijver9, K. Weck8, T. Hambuch4, T. Harkins5, D. Ballinger3, K. Voelkerding2, 11 1) Transformation Program Office, College of American Pathologists (CAP); 2) ARUP Laboratories; 3) Complete Genomics; 4) Illumina; 5) Life Technologies; 6) Department of Pathology, Stanford University School of Medicine; 7) Department of Laboratory Medicine & Pathology, Mayo Clinic Hospital; 8) Departments of Pathology & Laboratory Medicine and Genetics, University of North Carolina at Chapel Hill; 9) Stanford University School of Medicine, Departments of Pathology and Pediatrics; 10) Emory University School of Medicine; 11) Department of Pathology, University of Utah Department of Medicine; 12) Dept. Pathology, Brigham & Women's Hospital, Harvard Medical School; 13) Department of Pathology, University of Vermont College of Medicine; 14) UCLA; 15) University of Central Florida College of Medicine;; 16) Robert J Tomsich Pathology & Laboratory Medicine Institute; 17) Cleveland Clinic; 18) Cleveland Ohio.

   The higher throughput and lower cost of NGS has led to its rapid adoption in clinical testing. However, NGS-based tests are of much higher complexity than Sanger sequencing-based tests because of the requirements for extensive data analysis. In addition, there are challenges related to the acquisition and storage of data that far exceed those commonly generated in clinical laboratories. NGS tests are currently only offered as laboratory developed tests (LDTs) and require the development of clinical lab standards. The College of American Pathologists (CAP) recently developed accreditation requirements for labs offering NGS-based tests. Another mechanism of quality management mandated by CLIA is the requirement for routine proficiency testing. Here, we describe the development of a methods-based proficiency test (PT) developed at CAP that will be used to the monitor the proficiency of clinical labs offering NGS-based tests. The PT was designed to be suitable for all clinical labs using a variety of sequencing platforms and test applications and is appropriate for those offering either genome, exome, or gene panel tests. In order to create this PT product, we obtained genomic DNA from an established cell line derived from a healthy, appropriately consented individual. This CAP genome was sequenced at high depth by using three different technologies: Illumina (IL), Complete Genomics (CGI), and Ion Torrent (IT). There were 2 runs at the genome level using IL Hiseq (mean coverage [MC] 54X) and CGI (MC 44X) and 2 runs at the exome level using the IT PGM (MC 64X) at Life Technologies and IL HiSeq (MC 144X) at ARUP. The variants obtained using each centers internally developed pipeline are being analyzed for concordance to create a master list of variants that will consist of SNVs and indels. The pilot PT will query 200 concordant loci and will be a mix of SNVs, indels and wild type sequences. Participants will be sent genomic DNA to test their wet bench and bioinformatics pipelines and ability to correctly call variants. The result of our first pass analysis for characterizing the CAP genome for the 4 sequencing runs for exonic regions (coding, 5&3UTRs) showed 3827 (CGI), 1835 (IL-genome), 3495 (IL-exome), and 713 (IT-exome) SNVs that were unique for each run, with a total number of 21,575 concordant SNVs for all four runs. As expected, there is a lower concordance rate for indels. This methods-based PT will help ensure quality in clinical testing by NGS.

You may contact the first author (during and after the meeting) at