Detection of Copy Number Variants by Short Multiply Aggregated Sequence Homologies

Abstract

Chromosomal microarray testing is indicated for patients with diagnoses including unexplained developmental delay or intellectual disability, autism spectrum disorders, and multiple congenital anomalies. The short multiply aggregated sequence homologies (SMASH) genomic assay is a novel next-generation sequencing technology that performs copy number analysis at resolution similar to high-coverage whole genome sequencing but requires far less capacity. We benchmarked the performance of SMASH on a panel of genomic DNAs containing known copy number variants (CNVs). SMASH was able to detect pathogenic copy number variants of ≥10 kb in 77 of 77 samples. No pathogenic events were seen in 32 of 32 controls, indicating 100% sensitivity and specificity for detecting pathogenic CNVs >10 kb. Repeatability (interassay precision) and reproducibility (intra-assay precision) were assessed with 13 samples and showed perfect concordance. We also established that SMASH had a limit of detection of 20% for detection of large mosaic CNVs. Finally, we analyzed seven blinded specimens by SMASH analysis and successfully identified all pathogenic events. These results establish the efficacy of the SMASH genomic assay as a clinical test for the detection of pathogenic copy number variants at a resolution comparable to chromosomal microarray analysis.

Figure 1

Schematic of the short multiply aggregated sequence homologies (SMASH) Analytics version 1.0.1 pipeline (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY). The informatics processing steps following SMASH sequencing are detailed. SMASH utilizes a novel algorithm to parse minimal unique matches from chimeric sequence reads. These maps are then normalized and tabulated to generate a digital output of copy number across the genome. This workflow was developed using the Illumina HiSeq 2500 instrument, but can be readily adapted to any high-throughput sequencing technology. BCL, binary base call; GC correction, guanine and cytosine base composition normalization; QC, quality control; SA, SMASH analytics.

Supplemental Figure S1

Assay reproducibility (precision) schematic for short multiply aggregated sequence homologies reproducibility runs. Cross-hatching indicates different library preparations. For the interassay validation, the same pool of 13 samples was run on three different sequencing instruments on three separate days. Day indicates the time distribution of sequencing runs. Sequencers 1, 2, and 3 are three separate instruments; A and B are the two lanes of a HiSeq 2500 Rapid Run flow cell, only one of which was required for this study. The horizontal hatched lines represent the interassay test sample, which was loaded alone on day 1/sequencer 1 and day 3/sequencer 3 and multiplexed with the other intra-assay sample preparations on day 2/sequencer 2. For the intra-assay validation, 13 samples prepared in triplicate (represented by horizontal, vertical, and diagonal hatched lines) were run together on the same instrument on the same day (day 2/sequencer 2).