A benchmark of methods for SARS-CoV-2 whole genome sequencing and development of a more sensitive method

Since its inception in 2019, COVID-19 continues to cause infections and death, posing a significant health burden to the public. The causative agent of COVID-19 was discovered to be a virus that was named SARS-CoV-2. The genome (referring to the sum of all genes) of this virus is composed of 30,000 building blocks which scientists call bases. Soon after its discovery, several methods were developed to determine the entire sequence of these bases in what is termed whole genome sequencing (WGS). However, these different methods have different performance characteristics, which have not yet been adequately investigated. The challenge these methods face is that some samples contain very limited amounts of SARS-CoV-2 which makes it very challenging to accurately detect it. The other challenge is that the genes of SARS-CoV-2 change as the virus spreads from one person to another in a process called mutation. These mutations present a challenge to the different methods.

We set out to evaluate the performance of some of the commonly used methods for SARS-CoV-2 WGS. We specifically looked at how these methods would perform at varying levels of SARS-CoV-2 in a sample and how the methods would cope with SARS-CoV-2 genome harboring different mutations. We went further and devised a method that would be able to detect very small quantities of SARS-CoV-2 in a sample.

Our work has demonstrated that indeed, the methods developed for SARS-CoV-2 WGS have different performance characteristics. For example, a method termed ARTIC performed best in terms of multiplying the starting amount of SARS-CoV-2 in a sample. When we looked at varying levels of viral amounts in a sample, a method termed SNAP captured the largest portion of the genome when the virus amounts were abundant, whereas ARTIC would return the largest portion of the genome at high, medium, and low viral amounts of all the methods we tested. When it came to SARS-CoV-2 mutations, again the ARTIC method was the best in identifying the correct virus given the different mutations in samples at different levels of viral amounts. We concluded that, indeed, this ARTIC method is most well suited for SARS-CoV-2 WGS among the 7 methods we tested. We also report that we complemented this ARTIC method with another method called rolling-circle amplification to come up with a method that showed superior signs of detecting very minute amounts of SARS-CoV-2 in samples. Generally, our work has elaborated on the common methods used for SARS-CoV-2 WGS and attempted to devise a more sensitive method for SARS-CoV-2 WGS.

Read article