Deletion detection in SARS-CoV-2 genomes using multiplex-PCR sequencing from COVID-19 patients: elimination of false positives

We investigated sequencing and bioinformatic protocols for detecting deletions in SARS-CoV-2 genomes from COVID-19 patients. We adopted Illumina-based multiplex-PCR sequencing (specifically, ARTIC), the most widely used sequencing method for genetic characterization of SARS-CoV-2 in clinical samples. Available bioinformatics pipelines were tested with negative controls of human reference RNA with or without SARS-CoV-2 reference genome RNA; many false positive deletions were found. We optimized a bioinformatics pipeline to eliminate false positives based on features that differentiate them from true positives. We validated our optimized methods by sequencing and analyzing positive control samples (human reference RNA with synthetic SARS-CoV-2 RNA with deletion), and the known deletion was detected with few false positives.