Improving Diagnostic Success in Patients within the Undiagnosed Disease Network with Comprehensive Structural Variation Discovery Using Multiple Technologies

F. Sessions Cole, M.D., Daniel J. Wegner, M.S., and Jennifer A. Wambach, M.D., M.S.

The Undiagnosed Diseases Network (UDN) Clinical Site at Washington University School of Medicine (WUSM) evaluates patients with undiagnosed medical diseases who have been on prolonged diagnostic odysseys using careful clinical phenotyping and short read Illumina whole-genome sequencing (WGS) with structural variant (SV) detection using Manta. With these evaluations, the diagnostic success rate is ~35% (NEJM 2018;379:2131-9). Although SVs contribute substantially to genetic diversity and Mendelian disorders (Nature 2015;526:75; Genome Med 2018;10:95), comprehensive SV identification and characterization are difficult with short-read WGS strategies alone (PLoS Comput Biol 2015;11:e1004572; Bioinformatics 2009;25:2865; Nature 2015;517:608, Bioinformatics 2017;10:1093).

Using a multi-platform approach including short-read WGS with multiple SV detection algorithms, long-read WGS, strand-specific sequencing technologies, and optical mapping on 3 trios from the 1000 Genomes Project, Chaisson et al. were able to detect 3 to 7-fold more SVs than standard short-read sequencing alone (Nat Com 2019;10:1784:1-16).

We propose to test the hypothesis that state-of-the-art technologies including PacBio long-read WGS, BioNano optical mapping, and 10x phased WGS with and without de novo assembly, combined with reanalysis of UDN short-read WGS with multiple SV algorithms will identify pathogenic structural variants missed by standard UDN short-read WGS and improve diagnostic success among patients evaluated by the UDN clinical site at WUSM.

Print Friendly, PDF & Email