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

Whole-genome reconstruction of bacterial pathogens has become an important tool for tracking transmission and antimicrobial resistance gene spread, but highly accurate and complete assemblies have largely only historically been achievable using hybrid long- and short-read sequencing. We previously found the Oxford Nanopore Technologies (ONT) R10.4/kit12 flowcell/chemistry produced improved assemblies over the R9.4.1/kit10 combination, however long-read only assemblies contained more errors compared to Illumina-ONT hybrid assemblies. ONT have since released an R10.4.1/kit14 flowcell/chemistry upgrade and recommended the use of Bovine Serum Albumin (BSA) during library preparation, both of which reportedly increase accuracy and yield. They have also released updated basecallers trained using native bacterial DNA containing methylation sites intended to fix systematic basecalling errors, including common adenosine (A) to guanine (G) and cytosine (C) to thymine (T) substitutions. To evaluate these improvements, we successfully sequenced four bacterial reference strains, namely Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa and Staphylococcus aureus, and nine genetically diverse E. coli bloodstream infection-associated isolates from different phylogroups and sequence types, both with and without BSA. These sequences were de novo assembled and compared against Illumina-corrected reference genomes. In this small evaluation of 13 isolates we found that nanopore long-read-only R10.4.1/kit 14 assemblies with updated basecallers trained using bacterial methylated DNA produce accurate assemblies with ≥40×depth, sufficient to be cost-effective compared with hybrid ONT/Illumina sequencing in our setting.

Prediction of pyrazinamide resistance in Mycobacterium tuberculosis using structure-based machine-learning approaches.

Background: Pyrazinamide is one of four first-line antibiotics used to treat tuberculosis; however, antibiotic susceptibility testing for pyrazinamide is challenging. Resistance to pyrazinamide is primarily driven by genetic variation in pncA, encoding an enzyme that converts pyrazinamide into its active form. Methods: We curated a dataset of 664 non-redundant, missense amino acid mutations in PncA with associated high-confidence phenotypes from published studies and then trained three different machine-learning models to predict pyrazinamide resistance. All models had access to a range of protein structural-, chemical- and sequence-based features. Results: The best model, a gradient-boosted decision tree, achieved a sensitivity of 80.2% and a specificity of 76.9% on the hold-out test dataset. The clinical performance of the models was then estimated by predicting the binary pyrazinamide resistance phenotype of 4027 samples harbouring 367 unique missense mutations in pncA derived from 24 231 clinical isolates. Conclusions: This work demonstrates how machine learning can enhance the sensitivity/specificity of pyrazinamide resistance prediction in genetics-based clinical microbiology workflows, highlights novel mutations for future biochemical investigation, and is a proof of concept for using this approach in other drugs.

Rifampicin resistant Mycobacterium tuberculosis in Vietnam, 2020-2022.

Objective: We conducted a descriptive analysis of multi-drug resistant tuberculosis (MDR-TB) in Vietnam's two largest cities, Hanoi and Ho Chi Minh city. Methods: All patients with rifampicin resistant tuberculosis were recruited from Hanoi and surrounding provinces between 2020 and 2022. Additional patients were recruited from Ho Chi Minh city over the same time period. Demographic data were recorded from all patients, and samples collected, cultured, whole genome sequenced and analysed for drug resistance mutations. Genomic susceptibility predictions were made on the basis of the World Health Organization's catalogue of mutations in Mycobacterium tuberculosis associated with drug resistance, version 2. Comparisons were made against phenotypic drug susceptibility test results where these were available. Multivariable logistic regression was used to assess risk factors for previous episodes of tuberculosis. Results: 233/265 sequenced isolates were of sufficient quality for analysis, 146 (63 %) from Ho Chi Minh City and 87 (37 %) from Hanoi. 198 (85 %) were lineage 2, 20 (9 %) were lineage 4, and 15 (6 %) were lineage 1. 17/211 (8 %) for whom HIV status was known were infected, and 109/214 (51 %) patients had had a previous episode of tuberculosis. The main risk factor for a previous episode was HIV infection (odds ratio 5.1 (95 % confidence interval 1.3-20.0); p = 0.021). Sensitivity for predicting first-line drug resistance from whole genome sequencing data was over 90 %, with the exception of pyrazinamide (85 %). For moxifloxacin and amikacin it was 50 % or less. Among rifampicin-resistant isolates, prevalence of resistance to each non-first-line drug was < 20 %. Conclusions: Drug resistance among most MDR-TB strains in Vietnam's two largest cities is confined largely to first-line drugs. Living with HIV is the main risk factor among patients with MDR-TB for having had a previous episode of tuberculosis.

The plasmidome associated with Gram-negative bloodstream infections: A large-scale observational study using complete plasmid assemblies.

Plasmids carry genes conferring antimicrobial resistance and other clinically important traits, and contribute to the rapid dissemination of such genes. Previous studies using complete plasmid assemblies, which are essential for reliable inference, have been small and/or limited to plasmids carrying antimicrobial resistance genes (ARGs). In this study, we sequenced 1,880 complete plasmids from 738 isolates from bloodstream infections in Oxfordshire, UK. The bacteria had been originally isolated in 2009 (194 isolates) and 2018 (368 isolates), plus a stratified selection from intervening years (176 isolates). We demonstrate that plasmids are largely, but not entirely, constrained to a single host species, although there is substantial overlap between species of plasmid gene-repertoire. Most ARGs are carried by a relatively small number of plasmid groups with biological features that are predictable. Plasmids carrying ARGs (including those encoding carbapenemases) share a putative 'backbone' of core genes with those carrying no such genes. These findings suggest that future surveillance should, in addition to tracking plasmids currently associated with clinically important genes, focus on identifying and monitoring the dissemination of high-risk plasmid groups with the potential to rapidly acquire and disseminate these genes.

Hostile: accurate decontamination of microbial host sequences.

MOTIVATION: Microbial sequences generated from clinical samples are often contaminated with human host sequences that must be removed for ethical and legal reasons. Care must be taken to excise host sequences without inadvertently removing target microbial sequences to the detriment of downstream analyses such as variant calling and de novo assembly. RESULTS: To facilitate accurate host decontamination of both short and long sequencing reads, we developed Hostile, a tool capable of accurate host read removal using a laptop. We demonstrate that our approach removes at least 99.6% of real human reads and retains at least 99.989% of simulated bacterial reads. Using Hostile with a masked reference genome further increases bacterial read retention (≥99.997%) with negligible (≤0.001%) reduction in human read removal performance. Compared with an existing tool, Hostile removes 21%-23% more human short reads and 21-43 times fewer bacterial reads, typically in less time. AVAILABILITY AND IMPLEMENTATION: Hostile is implemented as an MIT-licensed Python package available from https://github.com/bede/hostile together with supplementary material.

Discordance between different bioinformatic methods for identifying resistance genes from short-read genomic data, with a focus on Escherichia coli.

Several bioinformatics genotyping algorithms are now commonly used to characterize antimicrobial resistance (AMR) gene profiles in whole-genome sequencing (WGS) data, with a view to understanding AMR epidemiology and developing resistance prediction workflows using WGS in clinical settings. Accurately evaluating AMR in Enterobacterales, particularly Escherichia coli, is of major importance, because this is a common pathogen. However, robust comparisons of different genotyping approaches on relevant simulated and large real-life WGS datasets are lacking. Here, we used both simulated datasets and a large set of real E. coli WGS data (n=1818 isolates) to systematically investigate genotyping methods in greater detail. Simulated constructs and real sequences were processed using four different bioinformatic programs (ABRicate, ARIBA, KmerResistance and SRST2, run with the ResFinder database) and their outputs compared. For simulation tests where 3079 AMR gene variants were inserted into random sequence constructs, KmerResistance was correct for 3076 (99.9 %) simulations, ABRicate for 3054 (99.2 %), ARIBA for 2783 (90.4 %) and SRST2 for 2108 (68.5 %). For simulation tests where two closely related gene variants were inserted into random sequence constructs, KmerResistance identified the correct alleles in 35 338/46 318 (76.3 %) simulations, ABRicate identified them in 11 842/46 318 (25.6 %) simulations, ARIBA identified them in 1679/46 318 (3.6 %) simulations and SRST2 identified them in 2000/46 318 (4.3 %) simulations. In real data, across all methods, 1392/1818 (76 %) isolates had discrepant allele calls for at least 1 gene. In addition to highlighting areas for improvement in challenging scenarios, (e.g. identification of AMR genes at <10× coverage, identifying multiple closely related AMR genes present in the same sample), our evaluations identified some more systematic errors that could be readily soluble, such as repeated misclassification (i.e. naming) of genes as shorter variants of the same gene present within the reference resistance gene database. Such naming errors accounted for at least 2530/4321 (59 %) of the discrepancies seen in real data. Moreover, many of the remaining discrepancies were likely 'artefactual', with reporting of cut-off differences accounting for at least 1430/4321 (33 %) discrepants. Whilst we found that comparing outputs generated by running multiple algorithms on the same dataset could identify and resolve these algorithmic artefacts, the results of our evaluations emphasize the need for developing new and more robust genotyping algorithms to further improve accuracy and performance.

Deep learning and single-cell phenotyping for rapid antimicrobial susceptibility detection in Escherichia coli.

The rise of antimicrobial resistance (AMR) is one of the greatest public health challenges, already causing up to 1.2 million deaths annually and rising. Current culture-based turnaround times for bacterial identification in clinical samples and antimicrobial susceptibility testing (AST) are typically 18-24 h. We present a novel proof-of-concept methodological advance in susceptibility testing based on the deep-learning of single-cell specific morphological phenotypes directly associated with antimicrobial susceptibility in Escherichia coli. Our models can reliably (80% single-cell accuracy) classify untreated and treated susceptible cells for a lab-reference fully susceptible E. coli strain, across four antibiotics (ciprofloxacin, gentamicin, rifampicin and co-amoxiclav). For ciprofloxacin, we demonstrate our models reveal significant (p < 0.001) differences between bacterial cell populations affected and unaffected by antibiotic treatment, and show that given treatment with a fixed concentration of 10 mg/L over 30 min these phenotypic effects correlate with clinical susceptibility defined by established clinical breakpoints. Deploying our approach on cell populations from six E. coli strains obtained from human bloodstream infections with varying degrees of ciprofloxacin resistance and treated with a range of ciprofloxacin concentrations, we show single-cell phenotyping has the potential to provide equivalent information to growth-based AST assays, but in as little as 30 min.

Persistence of SARS-CoV-2 antibodies over 18 months following infection: UK Biobank COVID-19 Serology Study.

BACKGROUND: Little is known about the persistence of antibodies after the first year following SARS-CoV-2 infection. We aimed to determine the proportion of individuals that maintain detectable levels of SARS-CoV-2 antibodies over an 18-month period following infection. METHODS: Population-based prospective study of 20 000 UK Biobank participants and their adult relatives recruited in May 2020. The proportion of SARS-CoV-2 cases testing positive for immunoglobulin G (IgG) antibodies against the spike protein (IgG-S), and the nucleocapsid protein (IgG-N), was calculated at varying intervals following infection. RESULTS: Overall, 20 195 participants were recruited. Their median age was 56 years (IQR 39-68), 56% were female and 88% were of white ethnicity. The proportion of SARS-CoV-2 cases with IgG-S antibodies following infection remained high (92%, 95% CI 90%-93%) at 6 months after infection. Levels of IgG-N antibodies following infection gradually decreased from 92% (95% CI 88%-95%) at 3 months to 72% (95% CI 70%-75%) at 18 months. There was no strong evidence of heterogeneity in antibody persistence by age, sex, ethnicity or socioeconomic deprivation. CONCLUSION: This study adds to the limited evidence on the long-term persistence of antibodies following SARS-CoV-2 infection, with likely implications for waning immunity following infection and the use of IgG-N in population surveys.

Evaluation of sequence hybridization for respiratory viruses using the Twist Bioscience Respiratory Virus Research panel and the OneCodex Respiratory Virus sequence analysis workflow.

Respiratory viral infections are a major global clinical problem, and rapid, cheap, scalable and agnostic diagnostic tests that capture genome-level information on viral variation are urgently needed. Metagenomic approaches would be ideal, but remain currently limited in that much of the genetic content in respiratory samples is human, and amplifying and sequencing the viral/pathogen component in an unbiased manner is challenging. PCR-based tests, including those which detect multiple pathogens, are already widely used, but do not capture information on strain-level variation; tests with larger viral repertoires are also expensive on a per-test basis. One intermediate approach is the use of large panels of viral probes or 'baits', which target or 'capture' sequences representing complete genomes amongst several different common viral pathogens; these are then amplified, sequenced and analysed with a sequence analysis workflow. Here we evaluate one such commercial bait capture method (the Twist Bioscience Respiratory Virus Research Panel) and sequence analysis workflow (OneCodex), using control (simulated) and patient samples head-to-head with a validated multiplex PCR clinical diagnostic test (BioFire FilmArray). We highlight the limited sensitivity and specificity of the joint Twist Bioscience/OneCodex approach, which are further reduced by shortening workflow times and increasing sample throughput to reduce per-sample costs. These issues with performance may be driven by aspects of both the laboratory (e.g. capacity to enrich for viruses present in low numbers), bioinformatics methods used (e.g. a limited viral reference database) and thresholds adopted for calling a virus as present or absent. As a result, this workflow would require further optimization prior to any implementation for respiratory virus characterization in a routine diagnostic healthcare setting.

Protection against SARS-CoV-2 Omicron BA.4/5 variant following booster vaccination or breakthrough infection in the UK.

Following primary SARS-CoV-2 vaccination, whether boosters or breakthrough infections provide greater protection against SARS-CoV-2 infection is incompletely understood. Here we investigated SARS-CoV-2 antibody correlates of protection against new Omicron BA.4/5 (re-)infections and anti-spike IgG antibody trajectories after a third/booster vaccination or breakthrough infection following second vaccination in 154,149 adults ≥18 y from the United Kingdom general population. Higher antibody levels were associated with increased protection against Omicron BA.4/5 infection and breakthrough infections were associated with higher levels of protection at any given antibody level than boosters. Breakthrough infections generated similar antibody levels to boosters, and the subsequent antibody declines were slightly slower than after boosters. Together our findings show breakthrough infection provides longer-lasting protection against further infections than booster vaccinations. Our findings, considered alongside the risks of severe infection and long-term consequences of infection, have important implications for vaccine policy.

Penicillin Binding Protein Substitutions Cooccur with Fluoroquinolone Resistance in Epidemic Lineages of Multidrug-Resistant Clostridioides difficile.

Clostridioides difficile remains a key cause of healthcare-associated infection, with multidrug-resistant (MDR) lineages causing high-mortality (≥20%) outbreaks. Cephalosporin treatment is a long-established risk factor, and antimicrobial stewardship is a key control. A mechanism underlying raised cephalosporin MICs has not been identified in C. difficile, but among other species, this is often acquired via amino acid substitutions in cell wall transpeptidases (penicillin binding proteins [PBPs]). Here, we investigated five C. difficile transpeptidases (PBP1 to PBP5) for recent substitutions, associated cephalosporin MICs, and co-occurrence with fluoroquinolone resistance. Previously published genome assemblies (n = 7,096) were obtained, representing 16 geographically widespread lineages, including healthcare-associated ST1(027). Recent amino acid substitutions were found within PBP1 (n = 50) and PBP3 (n = 48), ranging from 1 to 10 substitutions per genome. ß-Lactam MICs were measured for closely related pairs of wild-type and PBP-substituted isolates separated by 20 to 273 single nucleotide polymorphisms (SNPs). Recombination-corrected phylogenies were constructed to date substitution acquisition. Key substitutions such as PBP3 V497L and PBP1 T674I/N/V emerged independently across multiple lineages. They were associated with extremely high cephalosporin MICs; 1 to 4 doubling dilutions >wild-type, up to 1,506 µg/mL. Substitution patterns varied by lineage and clade, showed geographic structure, and occurred post-1990, coincident with the gyrA and/or gyrB substitutions conferring fluoroquinolone resistance. In conclusion, recent PBP1 and PBP3 substitutions are associated with raised cephalosporin MICs in C. difficile. Their co-occurrence with fluoroquinolone resistance hinders attempts to understand the relative importance of these drugs in the dissemination of epidemic lineages. Further controlled studies of cephalosporin and fluoroquinolone stewardship are needed to determine their relative effectiveness in outbreak control. IMPORTANCE Fluoroquinolone and cephalosporin use in healthcare settings has triggered outbreaks of high-mortality, multidrug-resistant C. difficile infection. Here, we identify a mechanism associated with raised cephalosporin MICs in C. difficile comprising amino acid substitutions in two cell wall transpeptidase enzymes (penicillin binding proteins). The higher the number of substitutions, the greater the impact on phenotype. Dated phylogenies revealed that substitutions associated with raised cephalosporin and fluoroquinolone MICs were co-acquired immediately before clinically important outbreak strains emerged. PBP substitutions were geographically structured within genetic lineages, suggesting adaptation to local antimicrobial prescribing. Antimicrobial stewardship of cephalosporins and fluoroquinolones is an effective means of C. difficile outbreak control. Genetic changes associated with raised MIC may impart a "fitness cost" after antibiotic withdrawal. Our study therefore identifies a mechanism that may explain the contribution of cephalosporin stewardship to resolving outbreak conditions. However, due to the co-occurrence of raised cephalosporin MICs and fluoroquinolone resistance, further work is needed to determine the relative importance of each.

A longitudinal study reveals persistence of antimicrobial resistance on livestock farms is not due to antimicrobial usage alone.

Introduction: There are concerns that antimicrobial usage (AMU) is driving an increase in multi-drug resistant (MDR) bacteria so treatment of microbial infections is becoming harder in humans and animals. The aim of this study was to evaluate factors, including usage, that affect antimicrobial resistance (AMR) on farm over time. Methods: A population of 14 cattle, sheep and pig farms within a defined area of England were sampled three times over a year to collect data on AMR in faecal Enterobacterales flora; AMU; and husbandry or management practices. Ten pooled samples were collected at each visit, with each comprising of 10 pinches of fresh faeces. Up to 14 isolates per visit were whole genome sequenced to determine presence of AMR genes. Results: Sheep farms had very low AMU in comparison to the other species and very few sheep isolates were genotypically resistant at any time point. AMR genes were detected persistently across pig farms at all visits, even on farms with low AMU, whereas AMR bacteria was consistently lower on cattle farms than pigs, even for those with comparably high AMU. MDR bacteria was also more commonly detected on pig farms than any other livestock species. Discussion: The results may be explained by a complex combination of factors on pig farms including historic AMU; co-selection of AMR bacteria; variation in amounts of antimicrobials used between visits; potential persistence in environmental reservoirs of AMR bacteria; or importation of pigs with AMR microbiota from supplying farms. Pig farms may also be at increased risk of AMR due to the greater use of oral routes of group antimicrobial treatment, which were less targeted than cattle treatments; the latter mostly administered to individual animals. Also, farms which exhibited either increasing or decreasing trends of AMR across the study did not have corresponding trends in their AMU. Therefore, our results suggest that factors other than AMU on individual farms are important for persistence of AMR bacteria on farms, which may be operating at the farm and livestock species level.

Enterobacterales plasmid sharing amongst human bloodstream infections, livestock, wastewater, and waterway niches in Oxfordshire, UK.

Plasmids enable the dissemination of antimicrobial resistance (AMR) in common Enterobacterales pathogens, representing a major public health challenge. However, the extent of plasmid sharing and evolution between Enterobacterales causing human infections and other niches remains unclear, including the emergence of resistance plasmids. Dense, unselected sampling is essential to developing our understanding of plasmid epidemiology and designing appropriate interventions to limit the emergence and dissemination of plasmid-associated AMR. We established a geographically and temporally restricted collection of human bloodstream infection (BSI)-associated, livestock-associated (cattle, pig, poultry, and sheep faeces, farm soils) and wastewater treatment work (WwTW)-associated (influent, effluent, waterways upstream/downstream of effluent outlets) Enterobacterales. Isolates were collected between 2008 and 2020 from sites <60 km apart in Oxfordshire, UK. Pangenome analysis of plasmid clusters revealed shared 'backbones', with phylogenies suggesting an intertwined ecology where well-conserved plasmid backbones carry diverse accessory functions, including AMR genes. Many plasmid 'backbones' were seen across species and niches, raising the possibility that plasmid movement between these followed by rapid accessory gene change could be relatively common. Overall, the signature of identical plasmid sharing is likely to be a highly transient one, implying that plasmid movement might be occurring at greater rates than previously estimated, raising a challenge for future genomic One Health studies.

Comparison of R9.4.1/Kit10 and R10/Kit12 Oxford Nanopore flowcells and chemistries in bacterial genome reconstruction.

Complete, accurate, cost-effective, and high-throughput reconstruction of bacterial genomes for large-scale genomic epidemiological studies is currently only possible with hybrid assembly, combining long- (typically using nanopore sequencing) and short-read (Illumina) datasets. Being able to use nanopore-only data would be a significant advance. Oxford Nanopore Technologies (ONT) have recently released a new flowcell (R10.4) and chemistry (Kit12), which reportedly generate per-read accuracies rivalling those of Illumina data. To evaluate this, we sequenced DNA extracts from four commonly studied bacterial pathogens, namely Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa and Staphylococcus aureus, using Illumina and ONT's R9.4.1/Kit10, R10.3/Kit12, R10.4/Kit12 flowcells/chemistries. We compared raw read accuracy and assembly accuracy for each modality, considering the impact of different nanopore basecalling models, commonly used assemblers, sequencing depth, and the use of duplex versus simplex reads. 'Super accuracy' (sup) basecalled R10.4 reads - in particular duplex reads - have high per-read accuracies and could be used to robustly reconstruct bacterial genomes without the use of Illumina data. However, the per-run yield of duplex reads generated in our hands with standard sequencing protocols was low (typically <10 %), with substantial implications for cost and throughput if relying on nanopore data only to enable bacterial genome reconstruction. In addition, recovery of small plasmids with the best-performing long-read assembler (Flye) was inconsistent. R10.4/Kit12 combined with sup basecalling holds promise as a singular sequencing technology in the reconstruction of commonly studied bacterial genomes, but hybrid assembly (Illumina+R9.4.1 hac) currently remains the highest throughput, most robust, and cost-effective approach to fully reconstruct these bacterial genomes.

Genomic Sequencing from Sputum for Tuberculosis Disease Diagnosis, Lineage Determination, and Drug Susceptibility Prediction.

Universal access to drug susceptibility testing for newly diagnosed tuberculosis patients is recommended. Access to culture-based diagnostics remains limited, and targeted molecular assays are vulnerable to emerging resistance mutations. Improved protocols for direct-from-sputum Mycobacterium tuberculosis sequencing would accelerate access to comprehensive drug susceptibility testing and molecular typing. We assessed a thermo-protection buffer-based direct-from-sample M. tuberculosis whole-genome sequencing protocol. We prospectively analyzed 60 acid-fast bacilli smear-positive clinical sputum samples in India and Madagascar. A diversity of semiquantitative smear positivity-level samples were included. Sequencing was performed using Illumina and MinION (monoplex and multiplex) technologies. We measured the impact of bacterial inoculum and sequencing platforms on genomic read depth, drug susceptibility prediction performance, and typing accuracy. M. tuberculosis was identified by direct sputum sequencing in 45/51 samples using Illumina, 34/38 were identified using MinION-monoplex sequencing, and 20/24 were identified using MinION-multiplex sequencing. The fraction of M. tuberculosis reads from MinION sequencing was lower than from Illumina, but monoplexing grade 3+ samples on MinION produced higher read depth than Illumina (P < 0.05) and MinION multiplexing (P < 0.01). No significant differences in sensitivity and specificity of drug susceptibility predictions were seen across sequencing modalities or within each technology when stratified by smear grade. Illumina sequencing from sputum accurately identified 1/8 (rifampin) and 6/12 (isoniazid) resistant samples, compared to 2/3 (rifampin) and 3/6 (isoniazid) accurately identified with Nanopore monoplex. Lineage agreement levels between direct and culture-based sequencing were 85% (MinION-monoplex), 88% (Illumina), and 100% (MinION-multiplex). M. tuberculosis direct-from-sample whole-genome sequencing remains challenging. Improved and affordable sample treatment protocols are needed prior to clinical deployment.

Virus Detection and Identification in Minutes Using Single-Particle Imaging and Deep Learning.

The increasing frequency and magnitude of viral outbreaks in recent decades, epitomized by the COVID-19 pandemic, has resulted in an urgent need for rapid and sensitive diagnostic methods. Here, we present a methodology for virus detection and identification that uses a convolutional neural network to distinguish between microscopy images of fluorescently labeled intact particles of different viruses. Our assay achieves labeling, imaging, and virus identification in less than 5 min and does not require any lysis, purification, or amplification steps. The trained neural network was able to differentiate SARS-CoV-2 from negative clinical samples, as well as from other common respiratory pathogens such as influenza and seasonal human coronaviruses. We were also able to differentiate closely related strains of influenza, as well as SARS-CoV-2 variants. Additional and novel pathogens can easily be incorporated into the test through software updates, offering the potential to rapidly utilize the technology in future infectious disease outbreaks or pandemics. Single-particle imaging combined with deep learning therefore offers a promising alternative to traditional viral diagnostic and genomic sequencing methods and has the potential for significant impact.

Evaluation of Nanopore sequencing for Mycobacterium tuberculosis drug susceptibility testing and outbreak investigation: a genomic analysis.

BACKGROUND: Mycobacterium tuberculosis whole-genome sequencing (WGS) has been widely used for genotypic drug susceptibility testing (DST) and outbreak investigation. For both applications, Illumina technology is used by most public health laboratories; however, Nanopore technology developed by Oxford Nanopore Technologies has not been thoroughly evaluated. The aim of this study was to determine whether Nanopore sequencing data can provide equivalent information to Illumina for transmission clustering and genotypic DST for M tuberculosis. METHODS: In this genomic analysis, we analysed 151 M tuberculosis isolates from Madagascar, South Africa, and England, which were collected between 2011 and 2018, using phenotypic DST and matched Illumina and Nanopore data. Illumina sequencing was done with the MiSeq, HiSeq 2500, or NextSeq500 platforms and Nanopore sequencing was done on the MinION or GridION platforms. Using highly reliable PacBio sequencing assemblies and pairwise distance correlation between Nanopore and Illumina data, we optimise Nanopore variant filters for detecting single-nucleotide polymorphisms (SNPs; using BCFtools software). We then used those SNPs to compare transmission clusters identified by Nanopore with the currently used UK Health Security Agency Illumina pipeline (COMPASS). We compared Illumina and Nanopore WGS-based DST predictions using the Mykrobe software and mutation catalogue. FINDINGS: The Nanopore BCFtools pipeline identified SNPs with a median precision of 99·3% (IQR 99·1-99·6) and recall of 90·2% (88·1-94·2) compared with a precision of 99·6% (99·4-99·7) and recall of 91·9% (87·6-98·6) using the Illumina COMPASS pipeline. Using a threshold of 12 SNPs for putative transmission clusters, Illumina identified 98 isolates as unrelated and 53 as belonging to 19 distinct clusters (size range 2-7). Nanopore reproduced 15 out of 19 clusters perfectly; two clusters were merged into one cluster, one cluster had a single sample missing, and one cluster had an additional sample adjoined. Illumina-based clusters were also closely replicated using a five SNP threshold and clustering accuracy was maintained using mixed Illumina and Nanopore datasets. Genotyping resistance variants with Nanopore was highly concordant with Illumina, having zero discordant SNPs across more than 3000 SNPs and four insertions or deletions (indels), across 60 000 indels. INTERPRETATION: Illumina and Nanopore technologies can be used independently or together by public health laboratories performing M tuberculosis genotypic DST and outbreak investigations. As a result, clinical and public health institutions making decisions on which sequencing technology to adopt for tuberculosis can base the choice on cost (which varies by country), batching, and turnaround time. FUNDING: Academy for Medical Sciences, Oxford Wellcome Institutional Strategic Support Fund, and the Swiss South Africa Joint Research Award (Swiss National Science Foundation and South African National Research Foundation).

Modern Solutions for Ancient Pathogens: Direct Pathogen Sequencing for Diagnosis of Lepromatous Leprosy and Cerebral Coenurosis.

Microbes unculturable in vitro remain diagnostically challenging, dependent historically on clinical findings, histology, or targeted molecular detection. We applied whole-genome sequencing directly from tissue to diagnose infections with mycobacteria (leprosy) and parasites (coenurosis). Direct pathogen DNA sequencing provides flexible solutions to diagnosis of difficult pathogens in diverse contexts.