Medizinische Universität Graz Austria/Österreich - Forschungsportal - Medical University of Graz
Gewählte Publikation:
Scheiber, MM.
Long-read sequencing based strategies for typing of Bordetella pertussis.
[ Diplomarbeit/Master Thesis (UNI) ] Universität Graz; 2021. pp.62.
FullText
- Autor*innen der Med Uni Graz:
- Betreuer*innen:
- Wagner-Lichtenegger Gabriel
- Altmetrics:
- Abstract:
- For outbreak tracing of B. pertussis, a slowly mutating but highly infectious pathogen, high-resolution typing on a genome level is advantageous. In a recent publication, Cabal et al. implemented a B. pertussis scheme for core genome MLST (cgMLST) to survey clusters and to monitor the genetic variability of B. pertussis isolates in Austria using short-read Illumina sequencing. Though high-throughput short-read sequencing produces vast amounts of NGS data, data acquisition time and cost can be obstacles. In contrast, Oxford Nanopore Technologies’ (ONT) MinION platform offers accessibility for a broader spectrum of researchers through low cost and portability, but was reported to be hampered by lower accuracy of the reads. Here, we investigated the applicability of nanopore sequencing on R10.3 Flow Cells for cgMLST, a method very sensitive to sequencing errors, using the B. pertussis isolates of the aforementioned study by Cabal et al. To counter the main disadvantage of ONT sequencing, the inherent high error rate of the raw reads, we evaluated several processing strategies, assemblers and polishing tools and aimed for a pipeline that generates assemblies as close as possible in accuracy to the one from Illumina short reads. The final pipeline is comprised of high accuracy basecalling, both-ends demultiplexing, assembly with Flye and polishing with Racon and Medaka, reducing the effect of sequencing errors to a maximum of six differing alleles compared to the short-read reference. On average 99.24 % of the scheme’s 2983 core genome targets were found in our assemblies. While the gained level of accuracy is not yet sufficient for the required discrimination depth to study closely related isolates, the methodology is perfectly suited for less detailed inquiries. Since ONT hardware and software are constantly improved, further pursuit of this approach holds great potential to advance global surveillance of bacterial pathogens.