Low-Cost DNA-Sequencing of Orphan Pediatric Retinal Diseases Using Custom Ampliseq Gene Panels.

Abstract

Purpose : In developing economies or those lacking universal health care, DNA sequencing is not available or not funded by health insurance. This is especially true for orphan retinal diseases. We wanted to develop rapid sequencing for eight genes involved in Norrie Disease, Familial Exudative Vitreo-Retinopathy (FEVR), and Retinoschisis, while reducing cost from several thousand dollars to less than $250 (US) per patient DNA sample. We also wanted an autonomous workflow that could be adopted by any regional university/medical-practice partnership, and which would also create new opportunities for medical genomic training of science and medical students.

Methods : DNA was extracted from 100 uL samples of whole frozen blood. An Ampliseq targeted-panel was designed using illumina's DesignStudio Sequencing Assay Designer for eight genes, 83 exons plus 25 bp adjacent intron padding. 180 amplicons in three pools. Target Genes: NDP, RS1 (Chr10); CTNNB1 (Chr3); TSPAN12 (Chr7); KIF11 (Chr10), FZD4, LRP5, ZNF408 (Chr11). Ampliseq libraries were analyzed by Agilent Bioanalyzer to confirm the average amplicon size (270 bp) plus adapters. Sequencing and variant calling was carried out on the iIlumina iSeq-100 platform, in our own lab, comparing pools of 16 and 39 samples for coverage.

Results : 2700-times amplicon coverage was obtained pooling 16 patient samples and 500-times coverage when pooling 39 patient samples. Numerous potential disease-associated variants were detected in targeted libraries from patients diagnosed with Norrie Disease, FEVR, and Retinoschisis. Average numbers of variant types per patient were: 20.1 ± 0.4 SNVs, 1.1 ± 0.3 Inserts, 0.76 ± 0.04 deletions.

Conclusions : We successfully developed a targeted gene sequencing panel that permits an autonomous in-house workflow from patient DNA to variant call analysis using the Illumina Ampliseq reagents and the iSeq-100 platform. The sequencing of 40 to 50 patients per run for eight genes costs less than $250 (US) per sample, including technician time, and provides for excellent base call accuracy (>Q30). Potential digenic variant contributions were detected, made possible by including multiple genetic components of the Norrin-FZD4 Wnt-signalling pathway. Our workflow can be adopted at even a small college or university and is suited for clinical practices with any inherited orphan disease patient population.

This is a 2020 ARVO Annual Meeting abstract.