Validation of publicly-available software used in analyzing NGS data for HIV-1 drug resistance mutations and transmission networks in a Washington, DC, Cohort

• Published in: PloS one 2019-04, Vol.14 (4), p.e0214820
• Main Authors: Jair, Kamwing, McCann, Chase D, Reed, Harrison, Castel, Amanda D, Pérez-Losada, Marcos, Wilbourn, Brittany, Greenberg, Alan E, Jordan, Jeanne A
• Format: Article
• Language: English
• Subjects: Adult; Anti-HIV Agents - therapeutic use; Bioinformatics; Biology and Life Sciences; Clusters; Cohort Studies; Computational biology; Computer programs; Correlation; Correlation analysis; Data Analysis; Data processing; Digital rights (Intellectual property); Disease transmission; District of Columbia; DNA polymerases; DNA sequencing; Drug resistance; Drug Resistance, Viral - drug effects; Drug Resistance, Viral - genetics; EDTA; Epidemiology; Evaluation; Female; Genetic aspects; GLP-1 receptor agonists; Health risks; High-Throughput Nucleotide Sequencing - methods; HIV; HIV Infections - drug therapy; HIV patients; HIV Seropositivity - drug therapy; HIV-1 - drug effects; HIV-1 - genetics; Homosexuality - drug effects; Human immunodeficiency virus; Humans; Integrase; Longitudinal Studies; Male; Medical laboratories; Medicine and Health Sciences; Methods; Middle Aged; Mutation; Mutation - genetics; Next-generation sequencing; Observational studies; Organic chemistry; Pediatrics; Proteases; Public health; Public relations; Public software; Research and Analysis Methods; Risk analysis; Risk factors; RNA-directed DNA polymerase; Software; Statistics; Viral Load - drug effects; Viral Load - genetics; Webs
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0214820

The DC Cohort is an ongoing longitudinal observational study of persons living with HIV. To better understand HIV-1 drug resistance and potential transmission clusters among these participants, we performed targeted, paired-end next-generation sequencing (NGS) of protease, reverse transcriptase and integrase amplicons. We elected to use free, publicly-available software (HyDRA Web, Stanford HIVdb and HIV-TRACE) for data analyses so that laboratory personnel without extensive bioinformatics expertise could use it; making the approach accessible and affordable for labs worldwide. With more laboratories transitioning away from Sanger-based chemistries to NGS platforms, lower frequency drug resistance mutations (DRMs) can be detected, yet their clinical relevance is uncertain. We looked at the impact choice in cutoff percentage had on number of DRMs detected and found an inverse correlation between the two. Longitudinal studies will be needed to determine whether low frequency DRMs are an early indicator of emerging resistance. We successfully validated this pipeline against a commercial pipeline, and another free, publicly-available pipeline. RT DRM results from HyDRA Web were compared to both SmartGene and PASeq Web; using the Mantel test, R2 values were 0.9332 (p