Optimization of the Illumina COVIDSeq™ protocol for decentralized, cost-effective genomic surveillance

Conclusion: COVID incident rate is sloping downward globally so is the financial support for high throughput genomic surveillance. However, persistent surveillance of circulating genomic variants in the community is crucial for scrutinizing the transmission dynamics of existing variants and tracking the emergence of new variants. We have optimized the SARS-CoV-2 variant detection assays for sequencing small batches (n = 48) of samples on a low-cost instrument (Illumina MiniSeq) which is ideal for small laboratories often not supported by centralized funding sources. Furthermore, surveillance data from such laboratories is critical for broadening the representation of the under-served population in global databases such as GSAID. Simultaneous sequencing of ~3000 samples using COVIDSeq™ EUA kit on Illumina NovaSeq (Price~$1.0 million + $100,000/ year maintenance cost) instrument is the most cost-effective (~$21.20/sample) option for SARS-CoV-2 genome sequencing often applied for mass surveillance. However, this published low cost is unattainable for independent clinical laboratories because of high capital investment and large batch size. Standard COVIDSeq™ protocol in Illumina’s EUA test allows the sequencing of up to 384 samples on the NextSeq 550 at a lower cost ($25.33/sample), which is slightly lower than the estimated cost ($43.27/sample) of this modified protocol for MinSeq. However, cost-effective testing will still require pooling of >300 samples to achieve ~$25/sample reagent cost. Capital investment for NextSeq 550 instrument (Cost ~$300,000 + ~$30,000/year maintenance cost) is still significantly higher than MiniSeq (Cost ~$50,000 + ~$5,000/year maintenance cost). Pre-pooling normalization has reduced the library preparation and sequencing cost on MiniSeq instruments close to the NextSeq or NovaSeq. Therefore, the modified protocol could empower small resource-limited laboratories to contribute to local genomic surveillance. We have adopted this modified protocol for sequencing 153 genomes from East Texas, USA, and compared the results with PCR-based variant detection [8]. High accuracy and reproducibility of this approach have been demonstrated in validating the COVIDSeq™ RUO assay for clinical application according to Clinical Laboratory Improvement Amendments (CLIA) and College of American Pathologists (CAP) guidelines [6]. We have only scrutinized this procedure for accurate detection of circulating variants rather than detecting new mutations and variant frequency in the mixed variant population. Secondly, the cost estimates presented in this study are for the core reagents (library preparation and sequencing) only. Personnel cost is a significant expense in processing NGS samples, which can vary significantly according to geographical location. For example, capital investment and the cost of imported reagents are often the limiting factors in low-income countries, not so much the trained personnel. Therefore, this cost-effective approach can still benefit the low-throughput sequencing for monitoring emerging variants of SARS-CoV-2 and support decentralized genomic surveillance, particularly in resource-limited settings.