Reliable Multiplex Sequencing with Rare Index Mis-Assignment on DNB-Based NGS Platform

Abstract

Accurate next generation sequencing (NGS) is critical for understanding genetic predisposition to human disease and thus aiding clinical diagnosis and personalized precision medicine. Recent breakthroughs in massively parallel sequencing, especially when coupled with sample multiplexing, have driven sequencing cost down and made clinical genetic tests broadly affordable. However, intractable index mis-assignment (commonly exceeds 1%) has been reported on some widely used sequencing platforms. Burdensome unique dual indexing is now used to reduce this problem. Here, we investigated this quality issue on BGI sequencers using three major library preparation methods: whole genome sequencing (WGS) with PCR, PCR-free WGS, and two-step targeted PCR. BGI's sequencers utilize a unique DNA nanoball (DNB) technology that is based on rolling circle replication (RCR) for array preparation; this linear amplification is PCR free and can avoid error accumulation. We demonstrate here that single index mis-assignment from free indexed oligos on these sequencers occurs at a rate of only one in 36 million reads, suggesting virtually no index hopping during DNB creation and arraying, as expected for the RCR process. Furthermore, the DNBbased NGS applications have achieved an unprecedentedly low sample-to-sample misassignment rate of 0.0001% to 0.0004% using only single indexing. Therefore, single indexing with DNB sequencing technology provides a simple but effective method for sensitive research and clinical genetic assays that require the detection of low abundance sequences in a large number of samples.

Prospects and challenges of implementing DNA metabarcoding for high-throughput insect surveillance

“…While a range of indexing strategies exist for HTS [153], for sensitive diagnostics applications it is critical to choose an approach that can adequately cope with the occasional recombination of these indices between molecules. Index-switching has received particular recent attention due to reports of remarkably high levels on current Illumina platforms [154]; however, similar phenomena can affect multiplexed sequencing across all major platforms to various degrees [155–159] (with the possible exception of recent MGI platforms [160]). Suggested causes include contamination from residual adapter/primer oligonucleotides [161], chimera formation during adapter PCR [162], mixed clusters on the flow cell [157], or physical contamination during library preparation or oligo synthesis by the vendor [159, 163, 164].…”

Multiple approaches for massively parallel sequencing of SARS-CoV-2 genomes directly from clinical samples

“…3b, c). Two reasons among others were likely to be accountable to this improvement: (1) we utilized 506 pieces of 120 ssDNA probes covering 2× of the SARS-CoV-2 genome to capture the libraries, and (2) we employed the DNBSEQ sequencing technology that features PCR-free rolling circle replication (RCR) of DNA nanoballs (DNBs) [41,42]. The sequencing results of amplicon and capture approaches revealed dramatic increases in the ratio of SARS-CoV-2 reads out of the total reads compared with meta sequencing, suggesting the enrichment was highly efficient-5596-fold in capture method and 5710-fold in amplicon method for each sample on average (Additional file 2 Table S2-S3).…”

“…1 The general workflow of multiple sequencing approaches adopted in this study. We employed unique dual indexing (UDI) strategy and DNB-based (DNA nanoball) PCR-free MPS platform to minimize index hopping and relevant sequencing errors [41][42][43]. a Amplicon-based enrichment: the UDI was integrated in the 2nd PCR.…”

Identification of viruses infecting six plum cultivars in Korea by RNA-sequencing

“…Multiple libraries are pooled into a single lane and sequenced by a single NGS platform. Index misassignments are frequently reported by some commercial NGS platforms ( Li et al, 2019 ). In fact, our six plum libraries were sequenced together with six peach libraries ( Jo et al, 2019 , 2018b ).…”