Misdetection of frameshifts in SARS-CoV-2 genomes: need for additional harmonisation and efficient monitoring of data workflows

Five years after the outbreak of the SARS-CoV-2 pandemic in 2020, diagnostic laboratories have moved from massive sequencing of thousands of samples to routine surveillance of SARS-CoV-2 cases, as with all other respiratory viruses. Surveillance remains of paramount importance to prevent a further SARS-CoV-2 surge, as the virus has been shown to mutate rapidly and can render available drugs and vaccines ineffective. During the pandemic, several bioinformatics pipelines and workflows have been developed to streamline analysis, shorten turnaround time and ensure reproducibility. As the number of samples decreases, laboratories are moving towards more flexible sequencing strategies and optimizing the cost per sample. However, workflow redesigns, even if individual steps have proven successful time and time again, can lead to challenges when changes in a bioinformatics pipeline are introduced (e.g. version updates, implementation of new features, etc.), a new combination of viral mutations emerge or a change in wet-lab procedures leads to unpredictable results. Here, we present a report of misidentified frameshift mutations in the consensus sequence of SARS-CoV-2, which led to an incorrect assumption of mutations in the spike and nucleocapsid viral proteins with the potential to affect PCR detection or even antigen testing. This investigation exemplifies the need for better awareness of the challenges that can occur even when using routinely applied protocols and analytical workflows and highlights the need for cooperation between experts of NGS, bioinformaticians and decision-makers towards more harmonized data workflows.