ReadItAndKeep: rapid decontamination of SARS-CoV-2 sequencing reads.

SUMMARY: Viral sequence data from clinical samples frequently contain contaminating human reads, which must be removed prior to sharing for legal and ethical reasons. To enable host read removal for SARS-CoV-2 sequencing data on low-specification laptops, we developed ReadItAndKeep, a fast lightweight tool for Illumina and nanopore data that only keeps reads matching the SARS-CoV-2 genome. Peak RAM usage is typically below 10 MB, and runtime less than 1 min. We show that by excluding the polyA tail from the viral reference, ReadItAndKeep prevents bleed-through of human reads, whereas mapping to the human genome lets some reads escape. We believe our test approach (including all possible reads from the human genome, human samples from each of the 26 populations in the 1000 genomes data and a diverse set of SARS-CoV-2 genomes) will also be useful for others. AVAILABILITY AND IMPLEMENTATION: ReadItAndKeep is implemented in C++, released under the MIT license, and available from https://github.com/GenomePathogenAnalysisService/read-it-and-keep. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Oral Contraceptives, Hormone Replacement Therapy, and Stroke Risk.

BACKGROUND: Millions of women worldwide use exogenous hormones as oral contraceptives or hormone replacement therapy. Still, time-dependent and long-term consequences of exogenous hormones on stroke risk remains unclear. METHODS: We examined the association between self-reported oral contraceptive and hormone replacement therapy use and stroke risk in 257 194 women from the UK Biobank, born between 1939 and 1970. Outcomes included any type of stroke, ischemic stroke, intracerebral hemorrhage, and subarachnoid hemorrhage. Exposures were analyzed as time-varying variables in Cox regression models. RESULTS: During first year of oral contraceptive use, an increased event rate of any stroke was observed (hazard ratio [HR], 2.49 [95% CI, 1.44-4.30]), while the hazards were found to be comparable during remaining years of use (HR, 1.00 [95% CI, 0.86-1.14]), compared with nonusers. Similarly, first year of hormone replacement therapy use was associated with higher hazard rates of any stroke (HR, 2.12 [95% CI, 1.66-2.70]), as well as cause-specific stroke, including ischemic stroke (HR, 1.93 [95% CI, 1.05-3.57]) and subarachnoid hemorrhage (HR, 2.17 [95% CI, 1.25-3.78]), which remained increased for any stroke during remaining years of use (HR, 1.18 [95% CI, 1.05-1.31]), and after discontinuation (HR, 1.16 [95% CI, 1.02-1.32]). CONCLUSIONS: Oral contraceptive use and hormone replacement therapy were associated with an increased risk of stroke, especially during the first year of use, possibly due to immediate changes in hemostatic balance. This study provides new insights on the effects of hormone exposure on stroke risk and provide evidence of not only an overall risk but also a pronounced effects seen in the beginning of treatment.

Predicting antibiotic resistance in complex protein targets using alchemical free energy methods.

Drug resistant Mycobacterium tuberculosis, which mostly results from single nucleotide polymorphisms in antibiotic target genes, poses a major threat to tuberculosis treatment outcomes. Relative binding free energy (RBFE) calculations can rapidly predict the effects of mutations, but this approach has not been tested on large, complex proteins. We use RBFE calculations to predict the effects of M. tuberculosis RNA polymerase and DNA gyrase mutations on rifampicin and moxifloxacin susceptibility respectively. These mutations encompass a range of amino acid substitutions with known effects and include large steric perturbations and charged moieties. We find that moderate numbers (n = 3-15) of short RBFE calculations can predict resistance in cases where the mutation results in a large change in the binding free energy. We show that the method lacks discrimination in cases with either a small change in energy or that involve charged amino acids, and we investigate how these calculation errors may be decreased.

Productivity in relation to organization of a surgical department: a retrospective observational study.

BACKGROUND: Responsible and efficient resource utilization are important factors in healthcare. The aim of this study was to investigate how total case time differs between two differently organized surgical departments. METHODS: This is a retrospective observational study of a cohort of patients undergoing elective surgery for breast cancer or malignant melanoma in a university hospital setting in Sweden. All patients were operated on by the same set of surgeons but in two different surgical departments: a general surgery (GS) and a cardiothoracic (CT) surgery department. Patients were selected to the two departments from a waiting list in the order of referral for surgery. The effect of being operated on at the CT department compared to the GS department was estimated by linear regression. RESULTS: The final study cohort comprised 349 patients in the GS department and 177 patients in the CT department. Both groups were similar regarding surgical procedures, American Society of Anesthesiologists' score, body mass index, age, sex, and the skill level of the operating surgeon. These covariates were included in the linear regression model. The total case time, defined by the Procedural Time Glossary as room set-up start to room clean-up finish, was significantly shorter for the patients who underwent a surgical procedure at the CT department compared to the GS department, even after adjusting for the background characteristics of the patients and surgeon. After adjusting for the selected covariates, the average difference in total case time between the two departments was - 30.67 min (p = 0.001). CONCLUSIONS: A significantly shorter total case time was measured for operations in the CT department. Plausible explanations may be more beneficial organizational factors, such as staffing ratio, skill mix in the operating room team, and working behavioral aspects regarding resource utilization.

Prediction of Susceptibility to First-Line Tuberculosis Drugs by DNA Sequencing.

BACKGROUND: The World Health Organization recommends drug-susceptibility testing of Mycobacterium tuberculosis complex for all patients with tuberculosis to guide treatment decisions and improve outcomes. Whether DNA sequencing can be used to accurately predict profiles of susceptibility to first-line antituberculosis drugs has not been clear. METHODS: We obtained whole-genome sequences and associated phenotypes of resistance or susceptibility to the first-line antituberculosis drugs isoniazid, rifampin, ethambutol, and pyrazinamide for isolates from 16 countries across six continents. For each isolate, mutations associated with drug resistance and drug susceptibility were identified across nine genes, and individual phenotypes were predicted unless mutations of unknown association were also present. To identify how whole-genome sequencing might direct first-line drug therapy, complete susceptibility profiles were predicted. These profiles were predicted to be susceptible to all four drugs (i.e., pansusceptible) if they were predicted to be susceptible to isoniazid and to the other drugs or if they contained mutations of unknown association in genes that affect susceptibility to the other drugs. We simulated the way in which the negative predictive value changed with the prevalence of drug resistance. RESULTS: A total of 10,209 isolates were analyzed. The largest proportion of phenotypes was predicted for rifampin (9660 [95.4%] of 10,130) and the smallest was predicted for ethambutol (8794 [89.8%] of 9794). Resistance to isoniazid, rifampin, ethambutol, and pyrazinamide was correctly predicted with 97.1%, 97.5%, 94.6%, and 91.3% sensitivity, respectively, and susceptibility to these drugs was correctly predicted with 99.0%, 98.8%, 93.6%, and 96.8% specificity. Of the 7516 isolates with complete phenotypic drug-susceptibility profiles, 5865 (78.0%) had complete genotypic predictions, among which 5250 profiles (89.5%) were correctly predicted. Among the 4037 phenotypic profiles that were predicted to be pansusceptible, 3952 (97.9%) were correctly predicted. CONCLUSIONS: Genotypic predictions of the susceptibility of M. tuberculosis to first-line drugs were found to be correlated with phenotypic susceptibility to these drugs. (Funded by the Bill and Melinda Gates Foundation and others.).

Reconciling the Potentially Irreconcilable? Genotypic and Phenotypic Amoxicillin-Clavulanate Resistance in Escherichia coli.

Resistance to amoxicillin-clavulanate, a widely used beta-lactam/beta-lactamase inhibitor combination antibiotic, is rising globally, and yet susceptibility testing remains challenging. To test whether whole-genome sequencing (WGS) could provide a more reliable assessment of susceptibility than traditional methods, we predicted resistance from WGS for 976 Escherichia coli bloodstream infection isolates from Oxfordshire, United Kingdom, comparing against phenotypes from the BD Phoenix (calibrated against EUCAST guidelines). A total of 339/976 (35%) isolates were amoxicillin-clavulanate resistant. Predictions based solely on beta-lactamase presence/absence performed poorly (sensitivity, 23% [78/339]) but improved when genetic features associated with penicillinase hyperproduction (e.g., promoter mutations and copy number estimates) were considered (sensitivity, 82% [277/339]; P < 0.0001). Most discrepancies occurred in isolates with MICs within ±1 doubling dilution of the breakpoint. We investigated two potential causes: the phenotypic reference and the binary resistant/susceptible classification. We performed reference standard, replicated phenotyping in a random stratified subsample of 261/976 (27%) isolates using agar dilution, following both EUCAST and CLSI guidelines, which use different clavulanate concentrations. As well as disagreeing with each other, neither agar dilution phenotype aligned perfectly with genetic features. A random-effects model investigating associations between genetic features and MICs showed that some genetic features had small, variable and additive effects, resulting in variable resistance classification. Using model fixed-effects to predict MICs for the non-agar dilution isolates, predicted MICs were in essential agreement (±1 doubling dilution) with observed (BD Phoenix) MICs for 691/715 (97%) isolates. This suggests amoxicillin-clavulanate resistance in E. coli is quantitative, rather than qualitative, explaining the poorly reproducible binary (resistant/susceptible) phenotypes and suboptimal concordance between different phenotypic methods and with WGS-based predictions.

Validating a 14-Drug Microtiter Plate Containing Bedaquiline and Delamanid for Large-Scale Research Susceptibility Testing of Mycobacterium tuberculosis.

The UKMYC5 plate is a 96-well microtiter plate designed by the CRyPTIC Consortium (Comprehensive Resistance Prediction for Tuberculosis: an International Consortium) to enable the measurement of MICs of 14 different antituberculosis (anti-TB) compounds for >30,000 clinical Mycobacterium tuberculosis isolates. Unlike the MYCOTB plate, on which the UKMYC5 plate is based, the UKMYC5 plate includes two new (bedaquiline and delamanid) and two repurposed (clofazimine and linezolid) compounds. UKMYC5 plates were tested by seven laboratories on four continents by use of a panel of 19 external quality assessment (EQA) strains, including H37Rv. To assess the optimal combination of reading method and incubation time, MICs were measured from each plate by two readers, using three methods (mirrored box, microscope, and Vizion digital viewing system), after 7, 10, 14, and 21 days of incubation. In addition, all EQA strains were subjected to whole-genome sequencing and phenotypically characterized by the 7H10/7H11 agar proportion method (APM) and by use of MGIT960 mycobacterial growth indicator tubes. We concluded that the UKMYC5 plate is optimally read using the Vizion system after 14 days of incubation, achieving an interreader agreement of 97.9% and intra- and interlaboratory reproducibility rates of 95.6% and 93.1%, respectively. The mirrored box had a similar reproducibility. Strains classified as resistant by APM, MGIT960, or the presence of mutations known to confer resistance consistently showed elevated MICs compared to those for strains classified as susceptible. Finally, the UKMYC5 plate records intermediate MICs for one strain for which the APM measured MICs close to the applied critical concentration, providing early evidence that the UKMYC5 plate can quantitatively measure the magnitude of resistance to anti-TB compounds that is due to specific genetic variation.

Automated detection of bacterial growth on 96-well plates for high-throughput drug susceptibility testing of Mycobacterium tuberculosis.

M. tuberculosis grows slowly and is challenging to work with experimentally compared with many other bacteria. Although microtitre plates have the potential to enable high-throughput phenotypic testing of M. tuberculosis, they can be difficult to read and interpret. Here we present a software package, the Automated Mycobacterial Growth Detection Algorithm (AMyGDA), that measures how much M. tuberculosis is growing in each well of a 96-well microtitre plate. The plate used here has serial dilutions of 14 anti-tuberculosis drugs, thereby permitting the MICs to be elucidated. The three participating laboratories each inoculated 38 96-well plates with 15 known M. tuberculosis strains (including the standard H37Rv reference strain) and, after 2 weeks' incubation, measured the MICs for all 14 drugs on each plate and took a photograph. By analysing the images, we demonstrate that AMyGDA is reproducible, and that the MICs measured are comparable to those measured by a laboratory scientist. The AMyGDA software will be used by the Comprehensive Resistance Prediction for Tuberculosis: an International Consortium (CRyPTIC) to measure the drug susceptibility profile of a large number (>30000) of samples of M. tuberculosis from patients over the next few years.

Effect of the Southeast Asian Ovalocytosis Deletion on the Conformational Dynamics of Signal-Anchor Transmembrane Segment 1 of Red Cell Anion Exchanger 1 (AE1, Band 3, or SLC4A1).

The first transmembrane (TM1) helix in the red cell anion exchanger (AE1, Band 3, or SLC4A1) acts as an internal signal anchor that binds the signal recognition particle and directs the nascent polypeptide chain to the endoplasmic reticulum (ER) membrane where it moves from the translocon laterally into the lipid bilayer. The sequence N-terminal to TM1 forms an amphipathic helix that lies at the membrane interface and is connected to TM1 by a bend at Pro403. Southeast Asian ovalocytosis (SAO) is a red cell abnormality caused by a nine-amino acid deletion (Ala400-Ala408) at the N-terminus of TM1. Here we demonstrate, by extensive (∼4.5 µs) molecular dynamics simulations of TM1 in a model 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine membrane, that the isolated TM1 peptide is highly dynamic and samples the structure of TM1 seen in the crystal structure of the membrane domain of AE1. The SAO deletion not only removes the proline-induced bend but also causes a "pulling in" of the part of the amphipathic helix into the hydrophobic phase of the bilayer, as well as the C-terminal of the peptide. The dynamics of the SAO peptide very infrequently resembles the structure of TM1 in AE1, demonstrating the disruptive effect the SAO deletion has on AE1 folding. These results provide a precise molecular view of the disposition and dynamics of wild-type and SAO TM1 in a lipid bilayer, an important early biosynthetic intermediate in the insertion of AE1 into the ER membrane, and extend earlier results of cell-free translation experiments.

Roles of Interleaflet Coupling and Hydrophobic Mismatch in Lipid Membrane Phase-Separation Kinetics.

Characterizing the nanoscale dynamic organization within lipid bilayer membranes is central to our understanding of cell membranes at a molecular level. We investigate phase separation and communication across leaflets in ternary lipid bilayers, including saturated lipids with between 12 and 20 carbons per tail. Coarse-grained molecular dynamics simulations reveal a novel two-step kinetics due to hydrophobic mismatch, in which the initial response of the apposed leaflets upon quenching is to increase local asymmetry (antiregistration), followed by dominance of symmetry (registration) as the bilayer equilibrates. Antiregistration can become thermodynamically preferred if domain size is restricted below ∼20 nm, with implications for the symmetry of rafts and nanoclusters in cell membranes, which have similar reported sizes. We relate our findings to theory derived from a semimicroscopic model in which the leaflets experience a "direct" area-dependent coupling, and an "indirect" coupling that arises from hydrophobic mismatch and is most important at domain boundaries. Registered phases differ in composition from antiregistered phases, consistent with a direct coupling between the leaflets. Increased hydrophobic mismatch purifies the phases, suggesting that it contributes to the molecule-level lipid immiscibility. Our results demonstrate an interplay of competing interleaflet couplings that affect phase compositions and kinetics, and lead to a length scale that can influence lateral and transverse bilayer organization within cells.

Alternating access mechanism in the POT family of oligopeptide transporters.

Short chain peptides are actively transported across membranes as an efficient route for dietary protein absorption and for maintaining cellular homeostasis. In mammals, peptide transport occurs via PepT1 and PepT2, which belong to the proton-dependent oligopeptide transporter, or POT family. The recent crystal structure of a bacterial POT transporter confirmed that they belong to the major facilitator superfamily of secondary active transporters. Despite the functional characterization of POT family members in bacteria, fungi and mammals, a detailed model for peptide recognition and transport remains unavailable. In this study, we report the 3.3-Å resolution crystal structure and functional characterization of a POT family transporter from the bacterium Streptococcus thermophilus. Crystallized in an inward open conformation the structure identifies a hinge-like movement within the C-terminal half of the transporter that facilitates opening of an intracellular gate controlling access to a central peptide-binding site. Our associated functional data support a model for peptide transport that highlights the importance of salt bridge interactions in orchestrating alternating access within the POT family.

Membrane stiffness is modified by integral membrane proteins.

The ease with which a cell membrane can bend and deform is important for a wide range of biological functions. Peripheral proteins that induce curvature in membranes (e.g. BAR domains) have been studied for a number of years. Little is known, however, about the effect of integral membrane proteins on the stiffness of a membrane (characterised by the bending rigidity, Kc). We demonstrate by computer simulation that adding integral membrane proteins at physiological densities alters the stiffness of the membrane. First we establish that the coarse-grained MARTINI forcefield is able to accurately reproduce the bending rigidity of a small patch of 1500 phosphatidyl choline lipids by comparing the calculated value to both experiment and an atomistic simulation of the same system. This enables us to simulate the dynamics of large (ca. 50 000 lipids) patches of membrane using the MARTINI coarse-grained description. We find that altering the lipid composition changes the bending rigidity. Adding integral membrane proteins to lipid bilayers also changes the bending rigidity, whilst adding a simple peripheral membrane protein has no effect. Our results suggest that integral membrane proteins can have different effects, and in the case of the bacterial outer membrane protein, BtuB, the greater the density of protein, the larger the reduction in stiffness.

Crystal structure of a prokaryotic homologue of the mammalian oligopeptide-proton symporters, PepT1 and PepT2.

PepT1 and PepT2 are major facilitator superfamily (MFS) transporters that utilize a proton gradient to drive the uptake of di- and tri-peptides in the small intestine and kidney, respectively. They are the major routes by which we absorb dietary nitrogen and many orally administered drugs. Here, we present the crystal structure of PepT(So), a functionally similar prokaryotic homologue of the mammalian peptide transporters from Shewanella oneidensis. This structure, refined using data up to 3.6 Å resolution, reveals a ligand-bound occluded state for the MFS and provides new insights into a general transport mechanism. We have located the peptide-binding site in a central hydrophilic cavity, which occludes a bound ligand from both sides of the membrane. Residues thought to be involved in proton coupling have also been identified near the extracellular gate of the cavity. Based on these findings and associated kinetic data, we propose that PepT(So) represents a sound model system for understanding mammalian peptide transport as catalysed by PepT1 and PepT2.

Compensatory mutations are associated with increased in vitro growth in resistant clinical samples of Mycobacterium tuberculosis.

Mutations in Mycobacterium tuberculosis associated with resistance to antibiotics often come with a fitness cost for the bacteria. Resistance to the first-line drug rifampicin leads to lower competitive fitness of M. tuberculosis populations when compared to susceptible populations. This fitness cost, introduced by resistance mutations in the RNA polymerase, can be alleviated by compensatory mutations (CMs) in other regions of the affected protein. CMs are of particular interest clinically since they could lock in resistance mutations, encouraging the spread of resistant strains worldwide. Here, we report the statistical inference of a comprehensive set of CMs in the RNA polymerase of M. tuberculosis, using over 70 000 M. tuberculosis genomes that were collated as part of the CRyPTIC project. The unprecedented size of this data set gave the statistical tests more power to investigate the association of putative CMs with resistance-conferring mutations. Overall, we propose 51 high-confidence CMs by means of statistical association testing and suggest hypotheses for how they exert their compensatory mechanism by mapping them onto the protein structure. In addition, we were able to show an association of CMs with higher in vitro growth densities, and hence presumably with higher fitness, in resistant samples in the more virulent M. tuberculosis lineage 2. Our results suggest the association of CM presence with significantly higher in vitro growth than for wild-type samples, although this association is confounded with lineage and sub-lineage affiliation. Our findings emphasize the integral role of CMs and lineage affiliation in resistance spread and increases the urgency of antibiotic stewardship, which implies accurate, cheap and widely accessible diagnostics for M. tuberculosis infections to not only improve patient outcomes but also prevent the spread of resistant strains.

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.

Addressing pandemic-wide systematic errors in the SARS-CoV-2 phylogeny.

The SARS-CoV-2 genome occupies a unique place in infection biology - it is the most highly sequenced genome on earth (making up over 20% of public sequencing datasets) with fine scale information on sampling date and geography, and has been subject to unprecedented intense analysis. As a result, these phylogenetic data are an incredibly valuable resource for science and public health. However, the vast majority of the data was sequenced by tiling amplicons across the full genome, with amplicon schemes that changed over the pandemic as mutations in the viral genome interacted with primer binding sites. In combination with the disparate set of genome assembly workflows and lack of consistent quality control (QC) processes, the current genomes have many systematic errors that have evolved with the virus and amplicon schemes. These errors have significant impacts on the phylogeny, and therefore over the last few years, many thousands of hours of researchers time has been spent in "eyeballing" trees, looking for artefacts, and then patching the tree. Given the huge value of this dataset, we therefore set out to reprocess the complete set of public raw sequence data in a rigorous amplicon-aware manner, and build a cleaner phylogeny. Here we provide a global tree of 3,960,704 samples, built from a consistently assembled set of high quality consensus sequences from all available public data as of March 2023, viewable at https://viridian.taxonium.org. Each genome was constructed using a novel assembly tool called Viridian (https://github.com/iqbal-lab-org/viridian), developed specifically to process amplicon sequence data, eliminating artefactual errors and mask the genome at low quality positions. We provide simulation and empirical validation of the methodology, and quantify the improvement in the phylogeny. Phase 2 of our project will address the fact that the data in the public archives is heavily geographically biased towards the Global North. We therefore have contributed new raw data to ENA/SRA from many countries including Ghana, Thailand, Laos, Sri Lanka, India, Argentina and Singapore. We will incorporate these, along with all public raw data submitted between March 2023 and the current day, into an updated set of assemblies, and phylogeny. We hope the tree, consensus sequences and Viridian will be a valuable resource for researchers.

Inclusion of minor alleles improves catalogue-based prediction of fluoroquinolone resistance in Mycobacterium tuberculosis.

Objectives: Fluoroquinolone resistance poses a threat to the successful treatment of tuberculosis. WGS, and the subsequent detection of catalogued resistance-associated mutations, offers an attractive solution to fluoroquinolone susceptibility testing but sensitivities are often less than 90%. We hypothesize that this is partly because the bioinformatic pipelines used usually mask the recognition of minor alleles that have been implicated in fluoroquinolone resistance. Methods: We analysed the Comprehensive Resistance Prediction for Tuberculosis: an International Consortium (CRyPTIC) dataset of globally diverse WGS Mycobacterium tuberculosis isolates, with matched MICs for two fluoroquinolone drugs and allowed putative minor alleles to contribute to resistance prediction. Results: Detecting minor alleles increased the sensitivity of WGS for moxifloxacin resistance prediction from 85.4% to 94.0%, without significantly reducing specificity. We also found no correlation between the proportion of an M. tuberculosis population containing a resistance-conferring allele and the magnitude of resistance. Conclusions: Together our results highlight the importance of detecting minor resistance-conferring alleles when using WGS, or indeed any sequencing-based approach, to diagnose fluoroquinolone resistance.

Bedaquiline and clofazimine resistance in Mycobacterium tuberculosis: an in-vitro and in-silico data analysis.

BACKGROUND: Bedaquiline is a core drug for the treatment of multidrug-resistant tuberculosis; however, the understanding of resistance mechanisms is poor, which is hampering rapid molecular diagnostics. Some bedaquiline-resistant mutants are also cross-resistant to clofazimine. To decipher bedaquiline and clofazimine resistance determinants, we combined experimental evolution, protein modelling, genome sequencing, and phenotypic data. METHODS: For this in-vitro and in-silico data analysis, we used a novel in-vitro evolutionary model using subinhibitory drug concentrations to select bedaquiline-resistant and clofazimine-resistant mutants. We determined bedaquiline and clofazimine minimum inhibitory concentrations and did Illumina and PacBio sequencing to characterise selected mutants and establish a mutation catalogue. This catalogue also includes phenotypic and genotypic data of a global collection of more than 14 000 clinical Mycobacterium tuberculosis complex isolates, and publicly available data. We investigated variants implicated in bedaquiline resistance by protein modelling and dynamic simulations. FINDINGS: We discerned 265 genomic variants implicated in bedaquiline resistance, with 250 (94%) variants affecting the transcriptional repressor (Rv0678) of the MmpS5-MmpL5 efflux system. We identified 40 new variants in vitro, and a new bedaquiline resistance mechanism caused by a large-scale genomic rearrangement. Additionally, we identified in vitro 15 (7%) of 208 mutations found in clinical bedaquiline-resistant isolates. From our in-vitro work, we detected 14 (16%) of 88 mutations so far identified as being associated with clofazimine resistance and also seen in clinically resistant strains, and catalogued 35 new mutations. Structural modelling of Rv0678 showed four major mechanisms of bedaquiline resistance: impaired DNA binding, reduction in protein stability, disruption of protein dimerisation, and alteration in affinity for its fatty acid ligand. INTERPRETATION: Our findings advance the understanding of drug resistance mechanisms in M tuberculosis complex strains. We have established an extended mutation catalogue, comprising variants implicated in resistance and susceptibility to bedaquiline and clofazimine. Our data emphasise that genotypic testing can delineate clinical isolates with borderline phenotypes, which is essential for the design of effective treatments. FUNDING: Leibniz ScienceCampus Evolutionary Medicine of the Lung, Deutsche Forschungsgemeinschaft, Research Training Group 2501 TransEvo, Rhodes Trust, Stanford University Medical Scientist Training Program, National Institute for Health and Care Research Oxford Biomedical Research Centre, Oxford University Hospitals NHS Foundation Trust, Bill & Melinda Gates Foundation, Wellcome Trust, and Marie Sklodowska-Curie Actions.

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.