Development and Comprehensive Evaluation of Culture-Independent, Long Amplicon-Based Targeted Next-Generation Sequencing Methods for Predicting Antimicrobial Resistance in Tuberculosis

The great variety of antimicrobial resistance (AMR) profiles among tuberculosis (TB) patients necessitates a comprehensive detection method. This study developed culture-independent, long amplicon-based targeted next-generation sequencing (tNGS) methods for predicting AMR across 16 drugs within the <i>Mycobacterium tuberculosis</i> complex (MTBC). Multiplex PCR amplification was employed to enrich 20 gene regions, with sequencing performed on either the Oxford Nanopore Technologies (ONT) or Illumina platforms. Customized bioinformatics pipelines provide a streamlined process from raw data to clinician-friendly reports. The ONT tNGS method has been optimized, and its performance has been thoroughly evaluated, utilizing Q20+ chemistry in combination with the R10.4.1 flow cell. It requires only 15 high-quality reads per target gene to accurately identify all variants, with the turnaround time taking 4 h and 50 min. Studies confirmed that this method effectively identifies <i>Mycobacterium</i> species and was highly resistant to interference from other clinical pathogens. To ensure optimal coverage, it is recommended to input at least 500 copies of the genome and sequence 500MB of high-quality FASTQ data. Diagnostic performance evaluations demonstrate that this method achieves 98.35% concordance with phenotypic drug susceptibility testing (pDST) and is consistent with the results obtained from Xpert MTB/RIF assays. The design of long amplicons not only ensures comprehensive coverage of target regions but also simplifies primer design, facilitating compatibility with various sequencing platforms. Compared with previous studies, the optimized ONT tNGS method in this study significantly improves turnaround time, detection accuracy, and the comprehensive coverage of mutations associated with AMR.