Impact of SARS-CoV-2 Mutations on PCR Assay Sequence Alignment

Abstract

Real-time reverse transcription polymerase chain reaction (RT-PCR) assays are the most widely used molecular tests for the detection of SARS-CoV-2 and diagnosis of COVID-19 in clinical samples. PCR assays target unique genomic RNA regions to identify SARS-CoV-2 with high sensitivity and specificity. In general, assay development incorporates the whole genome sequences available at design time to be inclusive of all target species and exclusive of near neighbors. However, rapid accumulation of mutations in viral genomes during sustained growth in the population can result in signature erosion and assay failures, creating situational blind spots during a pandemic. In this study, we analyzed the signatures of 43 PCR assays distributed across the genome against over 1.6 million SARS-CoV-2 sequences. We present evidence of significant signature erosion emerging in just two assays due to mutations, while adequate sequence identity was preserved in the other 41 assays. Failure of more than one assay against a given variant sequence was rare and mostly occurred in the two assays noted to have signature erosion. Assays tended to be designed in regions with statistically higher mutations rates. in silico analyses over time can provide insights into mutation trends and alert users to the emergence of novel variants that are present in the population at low proportions before they become dominant. Such routine assessment can also potentially highlight false negatives in test samples that may be indicative of mutations having functional consequences in the form of vaccine and therapeutic failures. This study highlights the importance of whole genome sequencing and expanded real-time monitoring of diagnostic PCR assays during a pandemic.

Keywords: COVID-19; PCR; RT-PCR; SARS-CoV-2; biosurveillance; diagnostics; pandemic; signature.

Figure 1

PSET assay definition. The assay definition includes square brackets and parentheses to delimit the primer sequences. Note that a probe is optional and that any sequence outside of the amplicon region is considered context for alignment purposes. This example corresponds to the cdc_n1 assay.

Figure 2

Genomic variation and assay regions. The numbered rectangles and labeled arrows indicate PCR assay and gene regions respectively. Heat values indicate the percentage of SNVs observed at the reference location across all subjects. Percentages for both SNVs and indels are compiled in Supplementary Table 4.

Figure 3

Regional mutation distributions. The two-sample Kolmogorov-Smirnov test rejected the null hypothesis (α = 0.05), supporting the elevated number of mutations observed in the assay regions (D = 0.408, p < 2.2e−16, alternative hypothesis: two-sided). Supplementary Tables 5, 6 list the counts by position and percentage of mutations by region.

Figure 4

Cumulative true positive rate. Vertical and horizontal facets divide the graph by Pango lineage and assay target gene. (A) The top line graph shows the logarithm of the cumulative total of subject sequences. (B) Heat map of the current PCR assays with the cumulative conditional true positive rate assessed from Apr 2020 to Feb 2021. The assay targeting specific genes are labeled on the right. White represents the absence of subject sequences of the lineage in the reference database. Vertical lines when each disease control organization called the lineage a VoC/VoI based on available data (compiled in Supplementary Table A).