Targeted-Sequencing Workflows for Comprehensive Drug Resistance Profiling of Mycobacterium tuberculosis Cultures Using Two Commercial Sequencing Platforms: Comparison of Analytical and Diagnostic Performance, Turnaround Time, and Cost

Abstract

Background The emergence of Mycobacterium tuberculosis with complex drug resistance profiles necessitates a rapid and comprehensive drug susceptibility test for guidance of patient treatment. We developed two targeted-sequencing workflows based on Illumina MiSeq and Nanopore MinION for the prediction of drug resistance in M. tuberculosis toward 12 antibiotics. Methods A total of 163 M. tuberculosis isolates collected from Hong Kong and Ethiopia were subjected to a multiplex PCR for simultaneous amplification of 19 drug resistance-associated genetic regions. The amplicons were then barcoded and sequenced in parallel on MiSeq and MinION in respective batch sizes of 24 and 12 samples. A web-based bioinformatics pipeline, BacterioChek-TB, was developed to translate the raw datasets into clinician-friendly reports. Results Both platforms successfully sequenced all samples with mean read depths of 1,127× and 1,649×, respectively. The variant calling by MiSeq and MinION could achieve 100% agreement if variants with an allele frequency of <40% reported by MinION were excluded. Both workflows achieved a mean clinical sensitivity of 94.8% and clinical specificity of 98.0% when compared with phenotypic drug susceptibility test (pDST). Turnaround times for the MiSeq and MinION workflows were 38 and 15 h, facilitating the delivery of treatment guidance at least 17–18 days earlier than pDST, respectively. The higher cost per sample on the MinION platform ($71.56) versus the MiSeq platform ($67.83) was attributed to differences in batching capabilities. Conclusion Our study demonstrates the interchangeability of MiSeq and MinION platforms for generation of accurate and actionable results for the treatment of tuberculosis.

Accuracy of an amplicon-sequencing nanopore approach to identify variants in tuberculosis drug-resistance-associated genes

“…Reducing the high number of FP variant calls required a high cut-off frequency (0.4) compared with Illumina (0.1), which compromised recall in our study. Other studies have also shown 100% of agreement in the prediction of drug-resistance-associated variants between Illumina and MinION Nanopore when variants detected by MinION at an allele frequency below 0.4 are excluded from the analysis [50,51]. Similarly, two recent studies showed complete concordance between both sequencing platforms when applied a threshold 0.8 to call fixed variants [52,53].…”

Nanopore Sequencing for Mycobacterium tuberculosis: a Critical Review of the Literature, New Developments, and Future Opportunities

“…We identified 58 articles of which 12 were eligible. Five articles focused on whole-genome sequencing (WGS) ( 8 , 12 , 21 – 23 ), three on targeted sequencing ( Table 1 ) ( 11 , 24 , 25 ), and four on software development for long-read ONT sequencing data with applications for M. tuberculosis ( Table 2 ) ( 16 , 18 , 19 , 26 ).…”

“…Three studies that assessed ONT sequencing for targeted sequencing of drug resistance loci were all published in 2020 ( 11 , 24 , 25 ). Tafess et al ( 24 ) and Chan et al ( 25 ) used custom-made panels of 19 and 10 drug resistance-associated loci, respectively.…”

Evaluation of whole-genome sequence data analysis approaches for short- and long-read sequencing of Mycobacterium tuberculosis

“…This compares to 100, 95, 82, 99 sensitivity, and 99, 100, 99, 99 % specificity, respectively, for the same drugs using the Mykrobe tool [12, 15]. However, there are limited numbers of studies using long-read (LR) sequencing data for anti-TB drug resistance prediction [9, 28] and for epidemiological typing [9].…”