Development of amplicon sequencing-based protocols for direct detection of antimicrobial resistance in mycobacterium tuberculosis and human immunodeficiency virus from clinical samples

Abstract

Direct sequencing of clinical specimens is now a trending approach for antibiotic resistance detection in target pathogenic microorganism as it can greatly reduce the time to report to a few working days that favors the choice of the appropriate regimens and the management of resource and samples. To further understand the process in developing sequencing workflows for this application, two target sequencing workflows: direct antimicrobial resistance (AMR) detection in Mycobacterium tuberculosis (MTB) in sputum samples and direct antiretroviral (ARV) resistance detection in human immunodeficiency virus 1 (HIV) in plasma samples, were successfully developed. The target sequencing workflow for AMR detection in MTB achieved 100% agreement between nanopore sequencing by Oxford Nanopore Technologies (ONT) and Illumina next generation sequencing (NGS) for the MTB DNA contents in samples above the limit of detection (LOD), while the target sequencing workflow for AVR resistance detection in HIV achieved high F1 score 0.918, or even 0.96 with a threshold from ROC analysis. The hierarchical clustering used in HIV sequencing workflow could even provide a detailed AVR resistance profile by associating the AVR resistance-associated amino acid mutation patterns with different quasispecies in the same samples. The success of these workflows proved the working principle of direct resistance detection in clinical specimens that requires only a few working days to report. Other than exploring the power of sequencing technologies, challenges for the workflow development were also highlighted. Both index misassignment and background nasal/oral flora (especially in sputum samples with low MTB gDNA content) can cause contamination that can lead to false results. The recommendations in this study including the choice of the index set, and the employment of decoy strategy can minimize the impact of these issues. The above findings can be a reference for the future drug resistance workflow development for other infectious diseases.