Abstract
This dataset comprises raw signal data from a multicenter study evaluating the accuracy of bacterial whole-genome genotyping using Oxford Nanopore long-read sequencing. The raw data comprises 79 isolates across six bacterial species, including 12 triplicates from three different laboratories (totalling ~1.4 TB of data). Sequencing was conducted on the latest R10.4.1 flow cells with V14 chemistry, producing on average 16 gigabases per flow cell. The generated raw ion current signals retain information beyond nucleotide sequences, supporting in-depth reanalysis for nucleotide modifications, resistance genes, and bacterial strain differentiation. The dataset enables re-basecalling with future models to keep up with the newest developments, e.g. to mitigate methylation-based calling errors, enhancing the reliability of SNP profiling and cgMLST analyses crucial for genomic surveillance. By sharing this raw signal data, accompanied by additional phenotypic resistance-data and an extensive quality control pipeline, we aim to advance reproducibility, support error correction studies and the continued development of bioinformatics tools, and encourage sharing raw data for broader genomic and epigenetic investigations as general best practice.
Subject terms: Bioinformatics, Bacterial genetics, Bacterial infection, Genetics research, DNA methylation
Background & Summary
Genomic surveillance of pathogens is crucial for tracking outbreaks1, monitoring antimicrobial resistance markers2, and informing public health interventions3,4. In the context of bacterial pathogens, nanopore sequencing has become a widely used tool for whole-genome sequencing and assembly, offering unprecedented accessibility through its portability, speed, and lower initial cost barriers while allowing researchers to obtain complete genomes with high accuracy and contiguity5–8. This capability is precious for applications in genomic surveillance, such as plasmid identification9,10, resistance gene annotation11, SNP profiling, and core genome multilocus sequence typing (cgMLST), the latter being critical for tracking bacterial outbreaks and understanding pathogen evolution in public health and surveillance contexts12,13. While the basecalled data in the form of FASTQ files, containing DNA sequences and the associated quality scores, is regularly shared among researchers on platforms such as the National Center for Biotechnology Information’s Sequence Read Archive14 (NCBI SRA) and the European Nucleotide Archive15 (EMBL-EBI ENA), raw signal data in the form of FAST5 or POD5 files (also called “squiggle” data) is rarely exchanged despite the advantages discussed below. The raw signal data may only be stored temporarily after basecalling, given the challenges of enormous file sizes, impracticalities in data upload, and the need for tailored data-sharing platforms that emphasize the importance of researchers also sharing their squiggles. Here, we share a comprehensive dataset of squiggle data from 79 bacterial isolates sequenced across three laboratories using Oxford Nanopore Technology (ONT). Our dataset enables researchers to explore signal-level analysis, develop improved algorithms, and optimize bioinformatic pipelines for applications beyond bacterial genotyping.
The complexity of the raw signal data is rooted in the nanopore sequencing mechanism, which generates long sequencing reads by measuring disturbances in the ion current as biological molecules such as DNA and RNA pass through a nanopore16,17. The recorded raw time-series current, the squiggle, is the unprocessed electrical signal output from ONT sequencing devices. During or after sequencing, the raw DNA/RNA signal can be transformed into nucleotide sequences through basecalling algorithms such as those implemented in ONT’s Dorado basecaller. A vital characteristic of the squiggle data is that it is shaped by any molecular feature that affects the flow of electric current through the nanopore and the speed at which the molecule moves through the pore (the translocation speed)16. Thus, nanopore sequencing can detect biological features beyond the mere stage of primary nucleotide composition in DNA by reading native RNA18, protein sequences19,20, various DNA modifications21 and RNA22 signals, such as those caused by methylation, secondary RNA structures23, and even whether a DNA belongs to a dead or living bacterium24. However, with all these technological possibilities comes a particular challenge of correctly translating the squiggle signal into the biological features of interest25,26.
The squiggle data represents a treasure trove of information that we are just beginning to utilize. Sharing raw signal data alongside FASTQ files also enhances the reproducibility of results. Furthermore, with the frequent updates to nanopore’s basecalling models, only access to the raw signal data allows for reanalysis using the latest and more accurate algorithms to improve overall read accuracy or to detect nucleotide modifications.
One specific example highlighting the importance of sharing squiggle data is methylation-based basecalling errors, which can impact bacterial genome reconstruction and subsequent genotyping. While nanopore sequencing has shown impressive capabilities across various applications, such strain-specific challenges related to miscalling methylated bases currently limit its applicability for high-resolution bacterial genotyping, as we and others have reported in recent studies26–29. Even a few incorrectly basecalled nucleotides can heavily impact the accuracy of SNP-based genotyping methods, which can lead to the misclassification of outbreaks in clinical settings. However, introducing the new v5 Dorado basecalling models, combined with a tailored Medaka v2 model30 trained specifically for bacterial methylation31, significantly enhanced genotyping accuracy, pointing toward a viable solution. In our multicenter performance study29, nanopore sequencing had already demonstrated robust and consistent results across participants with non-problematic strains; thus, further advances in basecalling approaches suggest it may also meet routine surveillance demands soon for problematic strains. In a follow-up study, we demonstrated that newer basecalling models reduce strand-specific ambiguity, although some errors persist for specific motifs32. Most importantly, re-assessing nanopore sequencing data in such a way, as described here but especially for future updates, is only possible if the raw signal data is shared.
Advancing beyond our initial investigation29 and the field’s standard practice of providing only base-called reads, we present comprehensive nanopore raw signal data that significantly extends our previous analysis of ONT accuracy and reproducibility in bacterial pathogen genotyping. This comprehensive whole-genome sequencing dataset offers unprecedented access to signal-level information, enabling e.g. enhanced methylation analysis and pathogen characterization previously unattainable through conventional approaches. We additionally include newly derived phenotypic metadata on the antimicrobial resistance profiles of all strains of clinically relevant species, which facilitate integrative analysis of phenotype and genotype. The raw data files shared here were sequenced by three laboratories using the latest R10.4.1 flow cells, V14 chemistry at the default translocation speed of 400 bp/s. In total, the shared data set comprises six publicly relevant bacterial species totalling 79 biosamples and 102 individual SRA runs based on the current default translocation speed (400 bp/s). This allows extensive re-analysis, for example, to assess error rates and to detect methylation signals. Here, we demonstrate how this data set can be re-basecalled and utilized to reduce methylation-induced errors in the generated assemblies, thus improving cgMLST analyses. We further demonstrate the reproducibility of methylation detection analysis in our dataset by focusing on triplicate samples, showing that similar methylation levels can be consistently detected in the same isolates when sequenced at different laboratories. At the same time, we also enable its use in training machine learning models, advancing algorithm development, and even supporting education in nanopore sequencing and methylation analysis, particularly given the limited availability of such open and comprehensive raw signal squiggle datasets.
Despite their importance for signal-level analyses amongst others, FAIR principles are often neglected for nanopore raw data due to data volume and lack of suitable repositories. The demand for appropriate infrastructure is reflected in the recent development of Squidbase (https://docs.squidbase.org/), specifically designed for nanopore squiggle data, alongside continued efforts to develop optimized storage formats such as SLOW533. With our comprehensive dataset including replicates, we make an important contribution toward making these valuable raw data comprehensively and sustainably available to the scientific community, contributing to the goal of a broader adoption of best practices for sharing nanopore raw data.
Methods
Strain selection and DNA isolation
We selected strains from six public-health-relevant bacterial species predominantly from Germany and Austria, collected between 2019–2022 from human hosts, to obtain a diverse collection of different sequence types29. These isolates comprise one Enterococcus faecalis (El), 19 Enterococcus faecium (Ef), 20 Klebsiella pneumoniae (Kp), 20 Listeria monocytogenes (Lm), 18 Staphylococcus aureus (Sa), and one Staphylococcus simulans (Ss) isolate, totaling in 79 different bacterial strains. A subset of 12 strains from four different species was selected based on previous results29 for sequencing in two additional laboratories to test the reproducibility of nanopore sequencing (Table 1). For these 12 strains, each laboratory had received identical pure cultures in stitch-agar propagated from a single colony in a blind-coded manner and had continued with cultivation and DNA preparation according to their protocols (Table 2). Since the study utilized only anonymized bacterial cultures with no connection to patient data and no additional specimens were collected beyond routine clinical care, the Ethics Committee of the University Medicine Greifswald confirmed that formal ethics approval was not required for this study.
Table 1.
Overview of cultivation, DNA preparation protocols and number of sequenced isolates per laboratory.
| Cultivation | DNA preparation | Number of sequenced isolates | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Medium | Duration | Temp. | Isolation | Single col. | El | Ef | Kp | Lm | Sa | Ss | |
| LAB1 | COL-S | 24 h | 37 °C | NucleoSpin + AXP Wash | yes (culture) | — | 3 | 3 | 3 | 3 | — |
| LAB2 | COL-S | 24 h | 37 °C | NucleoSpin + AXP Wash | yes (culture) | 1 | 19 | 20 | 20 | 18 | 1 |
| LAB3 | BHI | 16 h @ 160 rpm | 37 °C | MagAttract/ GenElute | yes (inoculation) | — | 3 | 3 | 3 | 2* | — |
The cultivation media were either solid Colombia sheep blood agar plates (COL-S) or liquid brain heart infusion (BHI) incubated at 37 °C overnight. The latter cultivation was performed in shakers for liquid media. Kits for DNA preparation/isolation were utilized according to the manufacturer’s instructions: NucleoSpin Microbial DNA (Macherey Nagel) and subsequent AXP Wash (Beckman Coulter) with a magnetic bead cleanup or MagAttract HMW DNA Kit Qiagen (in LAB3 for all species except Lm) and GenElute™ Bacterial Genomic DNA Kit (in LAB3 only for Lm). DNA isolation was performed from single-colony or liquid cultures. The species of the sequenced isolates are abbreviated as follows: El - Enterococcus faecalis, Ef - Enterococcus faecium, Kp - Klebsiella pneumoniae, Sa - Staphylococcus aureus, Ss - Staphylococcus simulans. 2*: Raw sequencing data for the third Sa isolate of LAB3 is unavailable due to unresolved technical issues during SRA deposition that persisted despite collaborative troubleshooting with NCBI support.
Table 2.
Summary of nanopore sequencing runs for the raw signal files uploaded in this data descriptor.
| Nanopore sequencing (SQK-NBD114.24, R10.4.1, 400 bp/s, 5 kHz) | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| DNA amount loaded (ng) | Samples per run | MinKNOW version | Average pores before sequencing | Number of used flow cells | Data produced (GB) | Bases sequenced (Gb) | N50 (kb) | Fail rate (%) | |
| LAB1 | 150 | 12 | 23.11.4 | 1140 | 1 | 174.98 | 15.05 | 12.6 | 18.4 |
| LAB2 | 90 | 16–20 | 23.04.5 | 1447 ± 73 | 5 | 997.23 | 58.27 | 8.3 ± 1.2 | 33.6 ± 8.7 |
| LAB3 | 82 | 12 | 23.07.12 | 1398 | 1 | 266.15 | 21.68 | 5.1 | 18.8 |
For LAB2, the values are either averages with standard deviations, or sums of all five runs (congruent with the number of used flow cells). The total of produced data per lab in gigabyte (GB) summarizes all files created during sequencing, including all POD5 files in passed and failed (and potentially skipped) folders, as well as basecalled FASTQ files and reports. The number of bases sequenced (in gigabases - Gb) correlates with the total data produced. Of all sequenced bases, a certain percentage is classified as failed based on, for example, low-quality base scores (noted in the fail rate).
The concentration of high-purity DNA from the respective preparation kits was measured with Qubit4 or Qubit Flex devices using the 1xds DNA BR kit or the 1xds DNA HS kit, depending on the expected DNA content.
Phenotypic antimicrobial resistance testing
The pure cultures of the strain collection of three clinically relevant species (Ef, Kp, Sa) were also analyzed in a fully automated VITEK® 2 XL instrument (bioMérieux, France) in LAB1 for phenotypic identification of antimicrobial susceptibility. Depending on the species the following card types were used, each incorporating 21 to 26 different antibiotic tests suitable for the respective species (AST-P655 for Enterococcus faecium, AST-N433 for Klebsiella pneumoniae, AST-P654 for Staphylococcus aureus) using the VITEK 2 Systems Version 9.03.3 and interpretation following the EUCAST 2023 guidelines.
Library preparation and nanopore sequencing
The library preparation was carried out in the same way in all laboratories using the ONT Native Barcoding Kit 24 V14 and following the protocol for gDNA ligation sequencing (SQK-NBD114.24). All sequencing was performed on R10.4.1 flow cells on MinION or GridION devices (Table 2). A set of 16–20 strains of the same species, or 12 strains from different species, to generate the sequencing triplicates, were sequenced on a single flow cell at the default translocation speed of 400 bp/s (5 kHz).
Contextualization of 400 bp/s signal data in relation to the previous 260 bp/s performance study
Building upon our previous performance study of 79 diverse bacterial strains sequenced across five laboratories29, we now provide the underlying raw nanopore signal data (squiggles). While FASTQ files were previously released (but based on now outdated basecalling models), the raw signal data presented here unlock advanced applications and (re-)analysis beyond traditional basecalling, including novel algorithm development and methylation analysis. Supplementary Table 1 provides comprehensive sequencing statistics that, while based on the original dataset, were exclusively compiled for this release, analogous to Table 2. This release provides, for the first time, access to raw squiggle data generated exclusively at the standard 400 bp/s translocation speed. Given that the legacy 260 bp/s mode has been deprecated and the resulting data are incompatible with contemporary basecalling algorithms and analytical pipelines, these obsolete raw signal data is excluded from this data descriptor. Of note, squiggle data is now typically saved as POD5 files during sequencing rather than in FAST5 format. However, we have to provide the raw signal data as basecalled FAST5 files due to the current limitations of SRA, which does not support POD5 data files (personal communication). Thus, the raw signal files must be subsequently converted to the original POD5 format and merged into a single file per barcode using a command from the POD5 file format software provided by ONT, as explained in detail in the Usage Notes section below.
Data Records
This data descriptor includes files representing direct whole-genome sequencing squiggle data of DNA isolations from pure bacterial cultures. The specific POD5/FAST5 file format contains the raw electrical signal measured in the nanopores with a recording speed of 400 bases per second sampled at 5 KHz. The main advantage of this raw data format is that the data is given the most significant possible reusability, as subsequent basecalling is possible with any available model suitable for the flow cell and translocation speed used. By default, this data is not often shared due to the large data storage requirements and the manual preparation required for a successful upload to common repositories. However, the availability of the raw squiggle data is the only way to keep the sequencing data up-to-date and usable for software updates.
For optimal data accessibility, Table 3 summarizes essential parameters for each squiggle data file provided with this data descriptor. For each isolate per laboratory, the table contains the library ID, i.e. the name by which individual files can be located online, the species, the size of each individual file, and the specific accession number.
Table 3.
Newly published squiggle data from six different bacterial species34.
| Library ID | Laboratory | Species | Size (GB) | SRA accession |
|---|---|---|---|---|
| EF21-PS_Lab2-raw | LAB2 | Enterococcus faecalis | 12 | SRR32250977 |
| EF22-PS_Lab1-raw | LAB1 | Enterococcus faecium | 16 | SRR31990277 |
| EF22-PS_Lab2-raw | LAB2 | Enterococcus faecium | 12 | SRR32250976 |
| EF22-PS_Lab3-raw | LAB3 | Enterococcus faecium | 1.8 | SRR31990273 |
| EF23-PS_Lab2-raw | LAB2 | Enterococcus faecium | 7.9 | SRR32250965 |
| EF24-PS_Lab2-raw | LAB2 | Enterococcus faecium | 13 | SRR32250954 |
| EF25-PS_Lab2-raw | LAB2 | Enterococcus faecium | 12 | SRR32250943 |
| EF26-PS_Lab1-raw | LAB1 | Enterococcus faecium | 12 | SRR31990276 |
| EF26-PS_Lab2-raw | LAB2 | Enterococcus faecium | 9.9 | SRR32250932 |
| EF26-PS_Lab3-raw | LAB3 | Enterococcus faecium | 4.1 | SRR31990272 |
| EF27-PS_Lab2-raw | LAB2 | Enterococcus faecium | 8.2 | SRR32250921 |
| EF28-PS_Lab2-raw | LAB2 | Enterococcus faecium | 9.8 | SRR32250910 |
| EF29-PS_Lab2-raw | LAB2 | Enterococcus faecium | 11 | SRR32250900 |
| EF30-PS_Lab2-raw | LAB2 | Enterococcus faecium | 5.9 | SRR32250899 |
| EF31-PS_Lab2-raw | LAB2 | Enterococcus faecium | 13 | SRR32250975 |
| EF32-PS_Lab2-raw | LAB2 | Enterococcus faecium | 6.4 | SRR32250974 |
| EF33-PS_Lab2-raw | LAB2 | Enterococcus faecium | 7.9 | SRR32250973 |
| EF34-PS_Lab2-raw | LAB2 | Enterococcus faecium | 12 | SRR32250972 |
| EF35-PS_Lab1-raw | LAB1 | Enterococcus faecium | 12 | SRR31990265 |
| EF35-PS_Lab2-raw | LAB2 | Enterococcus faecium | 13 | SRR32250971 |
| EF35-PS_Lab3-raw | LAB3 | Enterococcus faecium | 3.0 | SRR31990271 |
| EF36-PS_Lab2-raw | LAB2 | Enterococcus faecium | 7.1 | SRR32250970 |
| EF37-PS_Lab2-raw | LAB2 | Enterococcus faecium | 9.3 | SRR32250969 |
| EF38-PS_Lab2-raw | LAB2 | Enterococcus faecium | 11 | SRR32250968 |
| EF39-PS_Lab2-raw | LAB2 | Enterococcus faecium | 11 | SRR32250967 |
| EF40-PS_Lab2-raw | LAB2 | Enterococcus faecium | 11 | SRR32250966 |
| KP01-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 14 | SRR32250964 |
| KP02-PS_Lab1-raw | LAB1 | Klebsiella pneumoniae | 24 | SRR31990261 |
| KP02-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 15 | SRR32250963 |
| KP02-PS_Lab3-raw | LAB3 | Klebsiella pneumoniae | 14 | SRR31990270 |
| KP03-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 16 | SRR32250962 |
| KP04-PS_Lab1-raw | LAB1 | Klebsiella pneumoniae | 52 | SRR31990260 |
| KP04-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 19 | SRR32250961 |
| KP04-PS_Lab3-raw | LAB3 | Klebsiella pneumoniae | 13 | SRR31990269 |
| KP05-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 18 | SRR32250960 |
| KP06-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 27 | SRR32250959 |
| KP07-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 19 | SRR32250958 |
| KP08-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 24 | SRR32250957 |
| KP09-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 23 | SRR32250956 |
| KP10-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 27 | SRR32250955 |
| KP11-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 24 | SRR32250953 |
| KP12-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 17 | SRR32250952 |
| KP13-PS_Lab1-raw | LAB1 | Klebsiella pneumoniae | 39 | SRR31990259 |
| KP13-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 14 | SRR32250951 |
| KP13-PS_Lab3-raw | LAB3 | Klebsiella pneumoniae | 38 | SRR32190044 |
| KP14-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 19 | SRR32250950 |
| KP15-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 17 | SRR32250949 |
| KP16-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 15 | SRR32250948 |
| KP17-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 16 | SRR32250947 |
| KP18-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 17 | SRR32250946 |
| KP19-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 15 | SRR32250945 |
| KP20-PS_Lab2-raw | LAB2 | Klebsiella pneumoniae | 15 | SRR32250944 |
| LM41-PS_Lab1-raw | LAB1 | Listeria monocytogenes | 9.5 | SRR31990258 |
| LM41-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 14 | SRR32250942 |
| LM41-PS_Lab3-raw | LAB3 | Listeria monocytogenes | 34 | SRR31990267 |
| LM42-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 15 | SRR32250941 |
| LM43-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 11 | SRR32250940 |
| LM44-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 13 | SRR32250939 |
| LM45-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 11 | SRR32250938 |
| LM46-PS_Lab1-raw | LAB1 | Listeria monocytogenes | 11 | SRR31990257 |
| LM46-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 12 | SRR32250937 |
| LM46-PS_Lab3-raw | LAB3 | Listeria monocytogenes | 39 | SRR31990266 |
| LM47-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 9.5 | SRR32250936 |
| LM48-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 16 | SRR32250935 |
| LM49-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 11 | SRR32250934 |
| LM50-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 11 | SRR32250933 |
| LM51-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 11 | SRR32250931 |
| LM52-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 7.0 | SRR32250930 |
| LM53-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 13 | SRR32250929 |
| LM54-PS_Lab1-raw | LAB1 | Listeria monocytogenes | 13 | SRR31990256 |
| LM54-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 9.5 | SRR32250928 |
| LM54-PS_Lab3-raw | LAB3 | Listeria monocytogenes | 43 | SRR31990264 |
| LM55-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 8.3 | SRR32250927 |
| LM56-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 7.8 | SRR32250926 |
| LM57-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 7.3 | SRR32250925 |
| LM58-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 7.8 | SRR32250924 |
| LM59-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 7.5 | SRR32250923 |
| LM60-PS_Lab2-raw | LAB2 | Listeria monocytogenes | 12 | SRR32250922 |
| SA61-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 11 | SRR32250920 |
| SA62-PS_Lab1-raw | LAB1 | Staphylococcus aureus | 6.7 | SRR31990255 |
| SA62-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 2.7 | SRR32250919 |
| SA62-PS_Lab3-raw | LAB3 | Staphylococcus aureus | 7.0 | SRR31990263 |
| SA63-PS_Lab1-raw | LAB1 | Staphylococcus aureus | 4.5 | SRR31990275 |
| SA63-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 11 | SRR32250918 |
| SA63-PS_Lab3-raw | LAB3 | Staphylococcus aureus | 19 | SRR31990262 |
| SA65-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 12 | SRR32250917 |
| SA66-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 9.3 | SRR32250916 |
| SA67-PS_Lab1-raw | LAB1 | Staphylococcus aureus | 4.1 | SRR31990274 |
| SA67-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 12 | SRR32250915 |
| SA68-PS_Lab2-raw | LAB3 | Staphylococcus aureus | 11 | SRR32250914 |
| SA69-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 14 | SRR32250913 |
| SA70-PS_Lab2-raw | LAB2 | Staphylococcus simulans | 12 | SRR32250912 |
| SA71-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 13 | SRR32250911 |
| SA72-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 14 | SRR32250909 |
| SA73-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 9.4 | SRR32250908 |
| SA74-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 9.8 | SRR32250907 |
| SA75-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 11 | SRR32250906 |
| SA76-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 10 | SRR32250905 |
| SA77-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 7.3 | SRR32250904 |
| SA78-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 8.9 | SRR32250903 |
| SA79-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 7.9 | SRR32250902 |
| SA80-PS_Lab2-raw | LAB2 | Staphylococcus aureus | 6.9 | SRR32250901 |
Each row represents a squiggle data file (uploaded as FAST5) with its corresponding library ID, identifying the file in the SRA repository. The library ID is composed of: the abbreviated species, a sequential strain numbering, ‘PS’ for performance study, the laboratory that performed the sequencing (LAB1, LAB2, LAB3), and the suffix ‘-raw’, which indicates the raw squiggle data format. For each file, we also provide the file size in gigabytes and the respective SRA accession number, which serves as a unique identifier for downloading the file, for example via the following link (https://trace.ncbi.nlm.nih.gov/Traces/?view=run_browser&display=metada).
Important note on storage format and deposition: Although the temporary FAST5 storage format has now been largely replaced by POD5, all raw signal files had to be uploaded as basecalled FAST5 files according to the NCBI SRA rules and were added to the BioProject with accession number PRJNA109145234, also containing our previously published data29 for easy comparison. This was necessary because at the time of writing, neither SRA nor ENA accepted raw signal data in POD5 format (personal communication with the support). Our data descriptor includes scripts for the reconversion of the FAST5 files to state-of-the-art POD5 files (see below). The newly released squiggle data, provided in the temporary storage format FAST5), are labelled as “native ONT data from …” in the BioProject34. This distinguishes the current dataset from our previously published data in the same BioProject, where only basecalled FASTQ files were deposited, without the underlying raw signal data (squiggles) required for novel applications as presented here. For each bacterial isolate, two short-read data sets are also available from our previous study for further comparison of read accuracy29.
Relation with phenotypic results for antimicrobial resistance
The Vitek test results for the phenotypic antibiotic resistance of three species — Enterococcus faecium, Staphylococcus aureus, and Klebsiella pneumoniae — are presented in detailed tables at Zenodo35. The test results include the automatic interpretation from the Vitek system and the minimum inhibitory concentration (MIC) value for 21 to 26 different antibiotics being tested, depending on the species. This additional phenotypic metadata can enhance applications of our uploaded raw signal files towards comparison of genotype versus phenotype, resistance pattern analysis, and reproducibility of nanopore long-read sequencing.
Additional material35 was uploaded to Zenodo to enable comprehensive sharing of unfiltered data.
Technical Validation
We performed several validation steps to ensure the reliability and reproducibility of our raw signal data files, starting from the initial laboratory workflow related to our previous performance study, through read quality assessment, to a comprehensive validation using recognized genotyping methods in direct comparison to gold standard short-read sequencing data or between different basecalling pipelines. We further show the reproducibility of methylation detection throughout the available triplicates in our data set.
Laboratory workflow
The validation strategy leverages the multi-laboratory technical replication framework established in the previously published performance study, ensuring robust cross-laboratory reproducibility assessment. Identical pure cultures were distributed to different laboratories in blind-coded manner for unbiased processing using the same sequencing workflow, as described in the methods. Pre-sequencing quality control involved systematic DNA concentration measurement using Qubit4 or Qubit Flex devices coupled with 1xds DNA BR or 1xds DNA HS quantification kits to ensure adequate sample concentration for downstream sequencing. After sequencing, replicates from different laboratories yielded consistent species assignments and clustered accordingly after processing through our analysis pipelines, indicating technical accuracy.
Read statistics
The basic quality control of the basecalled reads from the released squiggle data files was carried out using NanoStat36 - as values reported here reflect basecalling outputs using current algorithms and may be refined through subsequent advances in basecalling software and computational models in the future. Table 4 summarizes the average values of all files per species and shows that the underlying squiggle data files are generally of good quality for nanopore reads. The quality of on average 98% of the raw reads per file is above Q10, and on average 68% of reads are above Q15, which can be significantly improved by further filtering, genome assembly, and its subsequent polishing. In addition, we achieved a high sequencing depth of 206-fold on average, as indicated by the number of bases per isolate.
Table 4.
Read statistics of basecalled squiggle data files of all three laboratories, averaged per species.
| Species | Av. N50 (bp) | Av. No. of bases per isolate | Av. Sequencing depth | Av read qual. (Q-score) |
|---|---|---|---|---|
| Enterococcus faecalis | 7063 | 615,435,097 | 228 | 14.1 |
| Enterococcus faecium | 8788 | 501,164,127 | 160 | 14.64 |
| Klebsiella pneumoniae | 9751 | 1,192,380,320 | 209 | 15.74 |
| Listeria monocytogenes | 8333 | 799,145,316 | 266 | 15.05 |
| Staphylococcus aureus | 8943 | 518,458,520 | 185 | 15.03 |
| Staphylococcus simulans | 11390 | 686,365,684 | 244 | 14.8 |
The average read N50 is typically around 10 kilobases for long reads and is higher for gram-negative bacteria, such as Klebsiella pneumoniae, due to less disruptive DNA extraction. The number of bases indicates a high average sequencing depth, exceeding 160 times the expected genome size per species. The read quality is noted as the Q-score, which can be translated to an error rate of about 0.05% for Q15.
Validation of long-read data in high-resolution genotyping in comparison to Illumina short-read data
Next, we demonstrate the recent improvements in nanopore data compared to Illumina short-read data in a high-resolution genotyping analysis, validating the usability of our provided raw read data (Fig. 1). Such analysis, which involves re-basecalling the ONT data with the latest tool versions and models, is only possible when the raw signal data are shared. Reanalyzing the ONT data in comparison to our previous performance study indeed reveals a reduction in potentially methylation-related basecalling errors in the final assemblies, as well as a higher concordance with the short-read-based assemblies.
Fig. 1.
cgMLST-based minimum spanning tree, analogous to the figures in Dabernig-Heinz et al.29. The tree includes 20 different Listeria monocytogenes isolates, indicating the minimum distance to the nearest neighbour on the connecting lines (number of differing cgMLST targets). For each isolate, two assemblies are included, generated from either short reads (SR, in green) or long reads (LR, white) of LAB2. Except for isolate LM46 (4 mismatching targets), there are no discrepancies between LR and SR. This represents a significant improvement over our previous analysis of the same raw read files with different basecalling and pipeline versions29, where LM46 had 66 mismatches, and again emphasizes the importance of sharing raw signal data for re-analysis with updated software.
The pipeline consists of basecalling with SUP version 5 models (Dorado version 8.3), assembly (flye 2.9.5), and polishing (medaka 2.0.1). With this evaluation of the raw signal data shared here, we can not only demonstrate correct species assignment, but also correct sequence types and comparable assemblies in major agreement with the short-read references. The assemblies of the replicates from different laboratories demonstrate reproducibility, thereby confirming the validation of the sequencing results (Table 5). Even for previously problematic bacterial isolates that exhibited substantial typing discrepancies compared to short-read data (four strains > = 3 mismatches, maximum of 66), newly basecalled and reanalyzed data markedly improved results (one strain > = 3 mismatches, maximum of 4), though minor inaccuracies persisted. The later underscores the ongoing requirement for basecalling algorithm advancement, for which raw signal data (squiggles), as demonstrated and provided in this study, will be essential for evaluation and development.
Table 5.
Genetic distance to the short-read reference of the long-read assemblies created in three different laboratories with the newest basecalling and polishing tools (dorado 0.9.0 and medaka 2.0.1).
| LAB1 | LAB2 | LAB3 | |
|---|---|---|---|
| EF22 | 1 | 0 | 1 |
| EF26 | 1 | 0 | 0 |
| EF35 | 0 | 0 | 0 |
| KP02 | 0 | 0 | 0 |
| KP04 | 0 | 0 | 0 |
| KP13 | 0 | 0 | 0 |
| LM41 | 0 | 0 | 0 |
| LM46 | 2 | 4 | 0 |
| LM54 | 0 | 0 | 0 |
| SA62 | 0 | 0 | 0 |
| SA63 | 0 | 0 | 1 |
| SA67 | 0 | 0 | 0 |
The typing results of mostly 0 and a maximum of 4 mismatches between the complete genome assemblies and the reference represent a significant improvement compared to previously published results (maximum of 66 mismatches), which were based on the same raw squiggle data, but basecalled and assembled with older software versions (dorado 0.4.0 and medaka 1.11.3). This clearly shows the benefit of ongoing efforts to improve basecalling accuracy and sharing raw signal squiggle data.
We applied the same evaluation pipeline to all other species (Fig. 2A–C) and the replicates from the three different laboratories (Table 5). Comparing the typing results between the replicates and the short-read reference led to 0–4 mismatches in the typing results, which represents a significant improvement compared to our previously published results, which had a maximum of 66 mismatches compared to the short reads of certain isolates29.
Fig. 2.
cgMLST-based minimum spanning trees, analogous to the figures in Dabernig-Heinz et al.29. These trees include both assemblies built from short reads (SR, in green) and from long reads from LAB2 (LR, white). (A) 19 Enterococcus faecium, (B) 20 different Klebsiella pneumoniae isolates and (C) 18 Staphylococcus aureus. There are no mismatches above the threshold of 3 between LR and the SR reference, except for isolate LM46 (4 mismatching targets to the reference). This is a significant improvement compared to our former analysis of the same raw read files with different basecalling versions and pipeline versions29.
Reproducibility of methylation and motif detection
The raw signal data from nanopore contains extensive epigenetic information, including methylation modifications such as 6 mA, 5mC, and 4mC, which are the most well-known modifications in bacterial epigenomes. To gain an overview of the methylation signals in our dataset, we ran Modkit37, a nanopore tool that processes raw reads containing methylation signals. After aligning the reads to a reference, the methylation level for each base (6 mA for adenine, and 5mC or 4mC for cytosine) is retrieved. We have developed and applied an in-house pipeline for preprocessing, base modification extraction, and motif detection38, focusing on the twelve strains with triplicate samples. Figure 3 summarizes the methylation percentages observed across replicates, showing high consistency in 6 mA detection relative to the total number of adenines. Additionally, several methylated motifs were identified with high confidence, with nearly all instances exhibiting full methylation (95–100%) in all replicates. A subset of motifs in L. monocytogenes was identified with only partial methylation throughout the genome (not all occurrences were methylated), which aligns with the higher fluctuations in 5mC and 4mC levels across replicates. Additionally, a specific motif in L. monocytogenes was found to be particularly challenging for the basecaller, as noted in both our study32 and another27, where this motif exhibited reduced accuracy in predicting methylated bases. This fluctuation suggests that potential biological or technical factors influencing methylation heterogeneity can be further investigated using our dataset. The complete list of methylated motifs and additional analyses can be found in Galeone et al.32, where we analyzed this dataset in greater detail regarding methylation.
Fig. 3.
Base methylation percentages across species with multiple replicates. The percentage of methylated bases is shown for 6 mA (number of methylated adenine over total number of adenine), 5mC (methylated cytosine over total cytosine), and 4mC (methylated cytosine over total cytosine) in species with multiple replicates across all three laboratories. Replicates generally agree on methylation levels, with strong consistency for 6 mA. In contrast, 5mC and 4mC exhibit more fluctuation, particularly in Listeria, likely due to specific motifs that appear to be partially methylated throughout the genome.
Overall, these results highlighted the reproducibility of methylation detection analysis across the replicates. The consistent identification of methylation patterns, particularly in 6 mA, demonstrates the robustness of this approach in capturing key epigenetic modifications.
Usage Notes
This section provides essential instructions for working with our nanopore raw signal data, including file download, conversion, basecalling, and methylation analysis.
File Download: Exemplary from SRA
The raw signal data files provided can be downloaded using the unique accession numbers in Table 3 to search for the respective experiment in a publicly available database such as SRA. A free account registration or login might be necessary to access this database. Following this link (https://trace.ncbi.nlm.nih.gov/Traces/?view=run_browser&display=metada), the SRA numbers can be copied into the Search field. By clicking the “Search” button, the respective page of the experiment will be opened, which includes useful metadata, automated analysis pipelines, read display, and, most importantly, the data access tab. Moving to this tab, the download link to the FAST5 file is shown under “Original format”. In most browsers, this link can be used for download by clicking on it, or the URL can be used to download with “wget” using a command line interface. Due to the extensive data format, which includes preliminary basecalls as well as the raw signal data that can be further processed with various current and future basecalling models, downloading the data requires substantial storage space (~1.4 TB for all 102 files) and a suitable download speed.
File Conversion: From FAST5 back to POD5
The raw data files were originally in POD5 format. Still, they had to be converted to the temporary storage format FAST5 for upload to the Sequence Read Archive, as SRA does not officially accept POD5 files yet (personal communication with the support). To work with the data using the latest basecalling models from ONT, the files need to be reconverted from FAST5 to POD5 after download.
To convert FAST5 files back to POD5 format, the pod5 package39 provided by ONT can be used. This can be done locally via the command line:
pod5 convert from_fast5 < fast5_file > -o < output_pod5_file > , or through the web-based interface available at https://pod5.nanoporetech.com.
Basecalling with Dorado
Once the data is converted back to POD5, we recommend using the Dorado basecaller for basecalling. A GPU is essential for optimal performance when using Dorado for basecalling, especially when using SUP models, as it significantly increases speed.
A minimal basecalling command is:
dorado basecaller sup < pod5_folder_path > > results.bam
This command will automatically select the latest basecalling models and operate in “super accuracy” (sup) mode for the highest accuracy. For more information about Dorado basecalling and its available options, refer to the Dorado GitHub page40.
If simultaneous methylation analysis is desired, Dorado also supports DNA modification basecalling. The following command can be used to call methylated bases, specifically 6 mA, 5mC, and 4mC modifications:
dorado basecaller sup,6 mA,4mC_5mC < pod5_folder_path > > results.bam
Methylation analysis pipeline
For downstream methylation analysis, we suggest using our methylation pipeline32, which processes the basecalled reads generated by Dorado, including the associated methylation calls. Specifically, it extracts methylated positions along a reference genome and identifies methylated motifs using Modkit. For further details on installing and using the pipeline, please refer to the user manual on GitHub37.
Supplementary information
Acknowledgements
We thank the participating institutions and colleagues from the former performance study, associated with the Austrian Agency for Health and Food Safety, the Institute for Infectious Diseases and Infection Control at Jena University Hospital, and the Robert Koch Institute (RKI) for their contributions to the publications. Their data allows for comprehensive comparability of the additional data reported here and places it in context to the gold standards of sequencing and previous ONT software versions. We thank the Digital Laboratory of the Genome Competence Center at the RKI for providing excellent data storage and delivery services, as well as technical support for obtaining the RKI raw signal data. DeepL and ChatGPT were used during the preparation of this manuscript for the sole purpose of better readability and language. We reviewed and edited the provided corrections and suggestions as needed and take full responsibility for the content of this publication. The authors state that no external funding was used.
Author contributions
Johanna Dabernig-Heinz: Laboratory Workflow, Data curation, Technical Validation, Visualization, Writing – original draft, review, and editing. Valentina Galeone: Laboratory Workflow, Data curation, Technical Validation, Visualization, Writing – original draft, review, and editing. Somayyeh Sedaghatjoo: Laboratory Workflow, Data curation, Technical Validation, Writing – review and editing. Ivo Steinmetz: Conceptualization, Methodology, Project administration, Resources, Supervision, Writing – review and editing. Christian Kohler: Laboratory Workflow, Data curation, Conceptualization, Methodology, Project administration, Writing – review and editing. Martin Hölzer: Conceptualization, Methodology, Data curation, Technical Validation, Visualization, Writing – original draft, review, and editing. Gabriel E. Wagner: Conceptualization, Methodology, Data curation, Technical Validation, Visualization, Project administration, Writing – original draft, review, and editing.
Funding
Open Access funding enabled and organized by Projekt DEAL.
Data availability
All newly published files contain direct whole-genome sequencing squiggle data in pod5/fast5 format for DNA isolations from 79 bacterial strains (Bioproject Information35). Table 3 contains the accession numbers and file size for each squiggle data file. These files have been added to BioProject PRJNA109145234, which also contains previously published data29. Further details on the bacterial strains in the BioProject, along with newly provided phenotypic AMR results, can be found in additional material on Zenodo35.
Code availability
No custom code was used in this study.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
These authors contributed equally: Johanna Dabernig-Heinz, Valentina Galeone.
Contributor Information
Martin Hölzer, Email: HoelzerM@rki.de.
Gabriel E. Wagner, Email: gabriel.wagner-lichtenegger@medunigraz.at
Supplementary information
The online version contains supplementary material available at 10.1038/s41597-025-06319-4.
References
- 1.Wyres, K. L., Lam, M. M. C. & Holt, K. E. Population genomics of Klebsiella pneumoniae. Nat. Rev. Microbiol.18, 344–359 (2020). [DOI] [PubMed] [Google Scholar]
- 2.Djordjevic, S. P. et al. Genomic surveillance for antimicrobial resistance — a One Health perspective. Nat. Rev. Genet.25, 142–157 (2024). [DOI] [PubMed] [Google Scholar]
- 3.Armstrong, G. L. et al. Pathogen Genomics in Public Health. N. Engl. J. Med.381, 2569–2580 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Gardy, J. L. & Loman, N. J. Towards a genomics-informed, real-time, global pathogen surveillance system. Nat. Rev. Genet.19, 9–20 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Bogaerts, B. et al. Closing the gap: Oxford Nanopore Technologies R10 sequencing allows comparable results to Illumina sequencing for SNP-based outbreak investigation of bacterial pathogens. J. Clin. Microbiol.62, e01576–23 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Wick, R. R., Judd, L. M. & Holt, K. E. Assembling the perfect bacterial genome using Oxford Nanopore and Illumina sequencing. PLOS Comput. Biol.19, e1010905 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Foster-Nyarko, E. et al. Nanopore-only assemblies for genomic surveillance of the global priority drug-resistant pathogen, Klebsiella pneumoniae. Microb. Genomics9, mgen000936 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Triebel, S. et al. De novo genome assembly resolving repetitive structures enables genomic analysis of 35 European Mycoplasmopsis bovis strains. BMC Genomics24, 548 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Zhao, W. et al. Oxford nanopore long-read sequencing enables the generation of complete bacterial and plasmid genomes without short-read sequencing. Front. Microbiol. 14 (2023). [DOI] [PMC free article] [PubMed]
- 10.Brown, S. D., Dreolini, L., Wilson, J. F., Balasundaram, M. & Holt, R. A. Complete sequence verification of plasmid DNA using the Oxford Nanopore Technologies’ MinION device. BMC Bioinformatics24, 116 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Sauerborn, E. et al. Detection of hidden antibiotic resistance through real-time genomics. Nat. Commun.15, 5494 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Struelens, M. J. et al. Real-time genomic surveillance for enhanced control of infectious diseases and antimicrobial resistance. Front. Sci. 2 (2024).
- 13.Werner, G. et al. Taking hospital pathogen surveillance to the next level. Microb. Genomics9, mgen001008 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Katz, K. et al. The Sequence Read Archive: a decade more of explosive growth. Nucleic Acids Res.50, D387–D390 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.O’Cathail, C. et al. The European Nucleotide Archive in 2024. Nucleic Acids Res.53, D49–D55 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Wang, Y., Zhao, Y., Bollas, A., Wang, Y. & Au, K. F. Nanopore sequencing technology, bioinformatics and applications. Nat. Biotechnol.39, 1348–1365 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Laszlo, A. H. et al. Decoding long nanopore sequencing reads of natural DNA. Nat. Biotechnol.32, 829–833 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Garalde, D. R. et al. Highly parallel direct RNA sequencing on an array of nanopores. Nat. Methods15, 201–206 (2018). [DOI] [PubMed] [Google Scholar]
- 19.Martin-Baniandres, P. et al. Enzyme-less nanopore detection of post-translational modifications within long polypeptides. Nat. Nanotechnol.18, 1335–1340 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Lu, C., Bonini, A., Viel, J. H. & Maglia, G. Toward single-molecule protein sequencing using nanopores. Nat. Biotechnol.43, 312–322 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Simpson, J. T. et al. Detecting DNA cytosine methylation using nanopore sequencing. Nat. Methods14, 407–410 (2017). [DOI] [PubMed] [Google Scholar]
- 22.Stephenson, W. et al. Direct detection of RNA modifications and structure using single-molecule nanopore sequencing. Cell Genomics2, 100097 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Bizuayehu, T. T. et al. Long-read single-molecule RNA structure sequencing using nanopore. Nucleic Acids Res.50, e120 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Urel, H. et al. Nanopore- and AI-empowered metagenomic viability inference. GigaScience. 14, giaf100 (2024). [DOI] [PMC free article] [PubMed]
- 25.Liu-Wei, W. et al. Sequencing accuracy and systematic errors of nanopore direct RNA sequencing. BMC Genomics25, 528 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Lohde, M. et al. Accurate bacterial outbreak tracing with Oxford Nanopore sequencing and reduction of methylation-induced errors. Genome Res.34, 2039–2047 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Biggel, M., Cernela, N., Horlbog, J. A. & Stephan, R. Oxford Nanopore's 2024 sequencing technology for Listeria monocytogenes outbreak detection and source attribution: progress and clone-specific challenges. J. Clin. Microbiol.62, e01083–24 (2024). [DOI] [PMC free article] [PubMed]
- 28.Linde, J. et al. Comparison of Illumina and Oxford Nanopore Technology for genome analysis of Francisella tularensis, Bacillus anthracis, and Brucella suis. BMC Genomics24, 258 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Dabernig-Heinz, J. et al. A multicenter study on accuracy and reproducibility of nanopore sequencing-based genotyping of bacterial pathogens. J. Clin. Microbiol.62, e00628–24 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.nanoporetech/medaka: Sequence correction provided by ONT Research. https://github.com/nanoporetech/medaka.
- 31.Wick, R. Medaka v2: progress and potential pitfalls. Ryan Wick's bioinformatics bloghttps://rrwick.github.io/2024/10/17/medaka-v2.html (2024).
- 32.Galeone, V. et al. Decoding bacterial methylomes in four public health-relevant microbial species: nanopore sequencing enables reproducible analysis of DNA modifications. BMC Genomics26, 394 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Samarakoon, H. et al. Flexible and efficient handling of nanopore sequencing signal data with slow5tools. Genome Biol.24, 69 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.NCBI Sequence Read Archivehttps://identifiers.org/ncbi/insdc.sra:SRP497546 (2024)
- 35.Dabernig-Heinz, J. A whole-genome sequencing dataset of nanopore raw signals for bacterial genotyping and methylation analysis - further information on AMR and BioProject. Zenodo10.5281/zenodo.17542640 [DOI] [PMC free article] [PubMed]
- 36.Lee, S. C.-H. & Burke, P. J. NanoStat: An open source, fully wireless potentiostat. Electrochimica Acta422, 140481 (2022). [Google Scholar]
- 37.nanoporetech/modkit. Oxford Nanopore Technologies (2025).
- 38.rki-mf1/ont-methylation. RKI MF1 Bioinformatics (2025).
- 39.nanoporetech/pod5-file-format. Oxford Nanopore Technologies (2025).
- 40.nanoporetech/dorado. Oxford Nanopore Technologies (2025).
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All newly published files contain direct whole-genome sequencing squiggle data in pod5/fast5 format for DNA isolations from 79 bacterial strains (Bioproject Information35). Table 3 contains the accession numbers and file size for each squiggle data file. These files have been added to BioProject PRJNA109145234, which also contains previously published data29. Further details on the bacterial strains in the BioProject, along with newly provided phenotypic AMR results, can be found in additional material on Zenodo35.
No custom code was used in this study.