20 July 2026 in SSRN Electronic Journal
BackgroundSwift tiled-amplicon and Arbor myBaits hybridization-capture protocols were widely used for SARS-CoV-2 whole-genome sequencing during the COVID-19 pandemic. Previous evaluations have focused on established lineages using contemporary reference databases and workflows. ObjectivesTo assess how enrichment strategy and bioinformatic analysis affect interpretation of a newly emerged SARS-CoV-2 lineage. MethodsWe analysed 30 Omicron-positive nasopharyngeal samples using Swift tiled-amplicon or Arbor myBaits hybridization-capture. Reads were processed with two nf-core-based workflows (covidseq and viralrecon), and lineage assignment performed using both contemporary and pre-Omicron reference resources. ResultsThe Swift protocol generated sequence data for all 30 samples, whereas the myBaits workflow failed for 12 samples with high Ct value. Among successfully sequenced samples, myBaits produced more uniform genome coverage, while Swift generally achieved higher median coverage. Bioinformatic workflow choice had limited impact on global sequencing metrics but influenced variant detection. With contemporary lineage definitions, all genomes were correctly classified as Omicron. Using historical pre-Omicron reference resources were used, lineage assignments differed substantially between enrichment approaches: myBaits consensus genomes were consistently assigned to ancestral B.1.1-derived lineages, whereas Swift-derived genomes were assigned to multiple lineages, including B.1.617.2 (Delta). These differences were associated with terminal SNVs, and modification of these regions altered historical lineage assignment. ConclusionsLineage assignment depends not only by reference databases and analytical workflows but also on which genomic regions are recovered. Although demonstrated using SARS-CoV-2, these results highlight a broader challenge for genomic surveillance systems confronted with previously unrecognised pathogen diversity.