SNES: single nucleus exome sequencing

Abstract

Single-cell genome sequencing methods are challenged by poor physical coverage and high error rates, making it difficult to distinguish real biological variants from technical artifacts. To address this problem, we developed a method called SNES that combines flow-sorting of single G1/0 or G2/M nuclei, time-limited multiple-displacement-amplification, exome capture, and next-generation sequencing to generate high coverage (96%) data from single human cells. We validated our method in a fibroblast cell line, and show low allelic dropout and false-positive error rates, resulting in high detection efficiencies for single nucleotide variants (92%) and indels (85%) in single cells.

Figure 1

SNES method and WGA quality control. (a) Nuclear suspensions were prepared from tissues, stained with DAPI and flow-sorted. Single nuclei were isolated by gating the G1/0 or G2/M ploidy distributions and deposited nuclei singly into a 96-well plate. Multiple-displacement-amplification is performed using Φ29 to perform WGA. (b) Time-course of WGA showing total DNA yield from single nuclei. (c) Quality control assay using a panel of 22 chromosome-specific qPCR primers to determine the WGA amplification efficiency of each single nucleus.

Figure 2

Coverage performance and metrics. (a) Coverage depth diagram. (b) Coverage depth data for G1/0 and G2/M single cells. (c) Coverage breadth diagram. (d) Coverage breadth data for exome region of G1/0 and G2/M single cells compared to previous studies using SNS [7] and MALBAC (Ni et al. [28]). Error bars show SEM. (e) Coverage uniformity diagram. (f) Coverage depth distribution for sites with low coverage in G1/0 and G2/M single cells. (g) Lorenz curves of coverage uniformity, showing values for perfect coverage, millions of SKN2 reference cells, Nuc-Seq single cell [11], single cells from G1/0 and G2/M distributions, and SNS cell [7]. (h) Capture performance of sequence reads across exons in the KRT76 locus for three single cells.

Figure 3

Error rates and detection efficiencies. (a) Illustration of technical errors, including the allelic-dropout rate (ADR) and false positive (FP) error rate. (b) Allelic dropout rates for single-cell experiments using G1/0 and G2/M single cells. (c) Spectrum of FP errors detected in the G2/M single-cell data, with each column representing a single cell. (d) Spectrum of single nucleotide variants detected in the SKN2 population data. (e) Distribution of FP errors on chromosome 7 and 11 that occur in single cells (black) or in two or more cells (red). (f) Detection efficiency for single-nucleotide variants in the G1/0 cells and G2/M cells calculated from the exome sequencing data. (g) Detection efficiency for indels in the G1/0 cells and G2/M cells calculated from single-cell exome sequencing data. Error bars in all panels represent the SEM.