Probe-Based Capture Enrichment Achieves 100% RSV Full-Genome Recovery

Respiratory syncytial virus (RSV) and human norovirus (HuNoV) together represent two of the most consequential viral pathogens globally — RSV as the leading cause of severe lower respiratory illness in infants and the elderly, and HuNoV as the dominant driver of acute gastroenteritis worldwide. Despite their clinical importance, obtaining high-quality, full-length genome sequences from clinical samples has remained difficult due to low viral titers, high host background, and the extraordinary genetic diversity across subtypes and genotypes. Existing commercial enrichment panels are designed for broad viral detection rather than deep, complete-genome recovery of these specific pathogens.

To address this gap, the Baylor College of Medicine team designed targeted oligonucleotide probe sets by mining 1,570 RSV and 1,376 HuNoV isolate sequences from GenBank, ensuring coverage of all major subtypes and genotypes. For RSV, the probe set was designed to capture both RSV-A and RSV-B subtypes, including the currently dominant ON and BA genotypes. For HuNoV, probes targeted GI.1, GII.4, GII.3, GII.6, and GII.17 genotypes. Libraries were prepared from RNA extracted from mid-turbinate nasal swabs (RSV) and stool samples or human intestinal enteroids (HuNoV), converted to cDNA, barcoded, pooled by Ct value, and sequenced on the Illumina NovaSeq 6000 platform at 2×150 bp.

The enrichment performance was striking. For RSV, the percentage of reads mapping to the viral genome rose from 0.08% pre-capture to 85.1% post-capture — more than a 1,000-fold enrichment. Average genome coverage increased from just 6× pre-capture to 123,524× post-capture (median). For HuNoV, viral reads increased from 1.15% to 40.8%, with average coverage rising from 2,285× to 241,333×. Critically, complete genomes (defined as >90% completeness, correct length range, and >20× coverage) were recovered from 100% of 85 RSV samples post-capture versus only 4% (1/24) pre-capture. For HuNoV, complete genomes were obtained in 85% (47/55) of samples post-capture versus 33% (18/55) pre-capture. Even samples with Ct values as high as 34.8 — near or below the detection limit — yielded complete or near-complete genomes after enrichment.

Phylogenetic and genotyping analyses confirmed accurate subtype and genotype assignment for all successfully assembled genomes, consistent with prior qPCR-based typing. RSV-A samples clustered within the ON genotype and RSV-B within the BA genotype, as expected for contemporary circulating strains. HuNoV genotyping correctly resolved GI.1, GII.4, and other GII genotypes. A particularly novel finding was the characterization of RSV subgenomic mRNA (sgmRNA) transcriptome patterns directly from clinical capture-enrichment data — the first such demonstration from patient specimens — revealing differential expression levels across the 10 RSV genes.

The study has important implications for viral genomic surveillance, vaccine development, and antiviral resistance monitoring. The ability to recover full-length genomes from low-titer clinical samples without amplicon-based approaches eliminates primer-binding bias and enables unbiased detection of novel variants. The workflow is scalable and compatible with standard Illumina infrastructure. Limitations include the relatively small sample set, the absence of samples with Ct values above 35 for systematic evaluation, and the fact that 8 of 55 HuNoV samples still failed to yield complete genomes post-capture, likely due to extremely low viral loads or degraded RNA.