Journal Article

P178 HIV-1 whole genome sequencing: comparative analysis of illumina and nanopore technologies

Bruzzone, B, Stefanelli, F, Ricucci, V, Taramasso, L, Orsi, A, Randazzo, N, Limongelli, A

2025

Overview

BackgroundWhole-genome sequencing (WGS) is a key tool for studying HIV-1 diversity, evolution, and epidemiology, especially as new drug classes emerge. The two main technologies, Illumina short-read and Oxford Nanopore (ONT) long-read sequencing, offer complementary advantages: Illumina’s high accuracy (<1% error) enables reliable variant detection, while ONT’s long reads improve haplotype resolution, essential for quasispecies analysis. However, despite improvements, ONT still has ~5% error rate, particularly in homopolymeric regions, which affects variant calling accuracy. This study evaluates WGS libraries obtained with AD4SEQ HIV-1 Whole Genome kit (Arrow Diagnostics) and sequenced on both technologies.Material and MethodsViral RNA was extracted using the ELITech InGenius® system. Illumina libraries were sequenced on iSeq100, while RT-PCR amplicons (~2600 bp) were processed with the Native Barcoding Kit 24 V14 (SQK-NBD114.24) and sequenced on MinION Mk1b for 19 h. Illumina FASTQ reads were analyzed using Smartvir (Smartseq) and an in-house pipeline, while ONT reads were processed with MinKnow and the EPI2ME pipeline. A 10% threshold consensus was obtained, focusing RAMs comparison initially on gag and pol, as they are easily analyzed on HIV Drug Resistance Database (Stanford University).ResultsOut of 24 samples, 2 were excluded due to library failure. The overall WGS coverage was similar for both methods as well as stratification for HIV-RNA levels (figure 1), except that ONT showed higher sample variability on low viral loads (table 1). Subtype-based analysis showed that subtype B had a mean WGS coverage of 83% for Illumina and ONT. Recombinant subtypes (CRF) exhibited the lowest coverage, with 57% (IQR: 45.0–69.0%) in Illumina and 66% (IQR: 48.8–82.1%) in ONT (figure 2). Gene-level analysis showed that at low viral loads (log 2–3), Illumina performed better in vif, tat, vpr and vpu, while ONT had superior coverage in nef; gag and pol were comparable. At higher viral loads, ONT showed better coverage for all accessory genes (figure 3). As per RAMs detection in pol, both technologies showed 100% concordance. In gag, 76% of mutations were detected by both platforms, 16% were found exclusively with ONT and 7% only with Illumina. Further investigations are ongoing.Abstract P178 Figure 1Overall WGS coverage in percentage compared between both sequencing technologies[Figure omitted. See PDF]Abstract P178 Figure 2Subtype-based stratification analysis of coverage between methods[Figure omitted. See PDF]Abstract P178 Figure 3Comparison of mean coverage (%). For simplicity only grouped log 2–3 and 6–7 are shown[Figure omitted. See PDF]Abstract P178 Table 1Stratified analysis of mean coverage in percentage per HIV-RNA levelConclusionsThese findings indicate that for pol RAM’s evaluation both technologies provide good results. As per coverage performance, Illumina ensures more stable sequencing for low viral loads, important for clinical samples, while Nanopore benefits from higher template availability, achieving superior coverage in high-viral-load samples. However, bioinformatics analysis parameters must also be carefully considered. While 99.9% accuracy is typical for Illumina (Q30), ONT operates around 90–95% accuracy (Q10), requiring additional filtering and error correction. Bioinformatics settings significantly impact sequencing outcomes and need to be further analyzed to refine analytical strategies.

• Publisher: BMJ Publishing Group LTD
• Subject: Accuracy / Bioinformatics / Comparative analysis / Drug resistance / Epidemiology / Genomes / HIV / Human immunodeficiency virus / Whole genome sequencing