A comparison of five Illumina, Ion Torrent, and nanopore sequencing technology-based approaches for whole genome sequencing of SARS-CoV-2

Rapid identification of the rise and spread of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants of concern remains critical for monitoring of the efficacy of diagnostics, therapeutics, vaccines, and control strategies. A wide range of SARS-CoV-2 next-generation sequencing (NGS) methods have been developed over the last years, but cross-sequence technology benchmarking studies have been scarce. In the current study, 26 clinical samples were sequenced using five protocols: AmpliSeq SARS-CoV-2 (Illumina), EasySeq RC-PCR SARS-CoV-2 (Illumina/NimaGen), Ion AmpliSeq SARS-CoV-2 (Thermo Fisher), custom primer sets (Oxford Nanopore Technologies (ONT)), and capture probe-based viral metagenomics (Roche/Illumina). Studied parameters included genome coverage, depth of coverage, amplicon distribution, and variant calling. The median SARS-CoV-2 genome coverage of samples with cycle threshold (Ct) values of 30 and lower ranged from 81.6 to 99.8% for, respectively, the ONT protocol and Illumina AmpliSeq protocol. Correlation of coverage with PCR Ct values varied per protocol. Amplicon distribution signatures differed across the methods, with peak differences of up to 4 log₁₀ at disbalanced positions in samples with high viral loads (Ct values ≤ 23). Phylogenetic analyses of consensus sequences showed clustering independent of the workflow used. The proportion of SARS-CoV-2 reads in relation to background sequences, as a (cost-)efficiency metric, was the highest for the EasySeq protocol. The hands-on time was the lowest when using EasySeq and ONT protocols, with the latter additionally having the shortest sequence runtime. In conclusion, the studied protocols differed on a variety of the studied metrics. This study provides data that assist laboratories when selecting protocols for their specific setting.