Evaluation of the accuracy of bacterial genome reconstruction with Oxford Nanopore R10.4.1 long-read-only sequencing

Abstract

2.Whole genome reconstruction of bacterial pathogens has become an import tool for tracking antimicrobial resistance spread, however accurate and complete assemblies have only been achievable using hybrid long and short-read sequencing. We have previously found the Oxford Nanopore Technologies (ONT) R10.4/kit12 flowcells produced improved assemblies over the R9.4.1/kit10, however they contained too many errors compared to hybrid Illumina-ONT assemblies. ONT have since released the R10.4.1/kit12 flowcells that promises greater accuracy and yield. They have also released newly trained basecallers using native bacterial DNA containing methylation sites intended to fix systematic errors, specifically Adenosine (A) to Guanine (G) and Cytosine (C) to Thymine (T) substitutions. ONT have recommended the use of Bovine Serum Albumin (BSA) during library preparation to improve sequencing yield and accuracy. To evaluate these improvements, we sequenced DNA extracts from four commonly studied bacterial pathogens, namely Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa and Staphylococcus aureus, as well as 12 disparate E. coli clinical samples from different phylogroups and sequence types. These were all sequenced with and without BSA. These sequences were de novo assembled and compared against Illumina-corrected reference genomes. Here we have found the nanopore long read-only R10.4.1 (kit14) assemblies with basecallers trained using native bacterial methylated DNA produce accurate assemblies from 40x depth or higher, sufficient to be cost-effective compared to hybrid long-read (ONT) and short-read (Illumina) sequencing. 3. Impact statementCurrently, the best method of building accurate and complete bacterial genome assemblies is to create a hybrid assembly; combining both long and short DNA sequencing reads. Short reads are more accurate, but can be difficult to assemble into a complete genome. Long reads are generally less accurate, but easier to reconstruct into a complete genome. By combining long and short reads, we get both accuracy and reconstructive power. However, this also involves higher costs and more labour than using a single sequencing platform. In this study, we compare long read only assemblies from Oxford Nanopore Technologys newest iteration of improvements in both chemistry and software to hybrid Illumina-Nanopore assemblies. We sequenced four bacterial pathogens with published reference genomes (Staphylococcus aureus, Klebsiella Pneumoniae, Pseudomonas Aeruginosa, and Escherichia Coli) and twelve bloodstream associated E. coli, and show that assemblies from the newest technology are not only an improvement on the previous iteration, but are able to compete with hybrid Illumina-Nanopore assemblies in their quality, providing a step towards bacterial genome assembly using a single sequencing platform. 4. Data summaryThe authors confirm all supporting data, code and protocols have been provided within the article, through supplementary data files, or in publicly accessible repositories. Nanopore and Illumina fastq data are available in the ENA under project accession: PRJEB51164. Assemblies have been made available at: https://figshare.com/articles/dataset/R10_4_1_KIT14_comparison_assemblies/2497 2954 Code and analysis outputs are available at: https://gitlab.com/ModernisingMedicalMicrobiology/assembly_comparison