| 1. |
|
|
| 2. |
- Benkwitz-Bedford, Sam, et al.
(författare)
-
Machine Learning Prediction of Resistance to Subinhibitory Antimicrobial Concentrations from Escherichia coli Genomes.
- 2021
-
Ingår i: mSystems. - 2379-5077. ; 6:4
-
Tidskriftsartikel (refereegranskat)abstract
- Escherichia coli is an important cause of bacterial infections worldwide, with multidrug-resistant strains incurring substantial costs on human lives. Besides therapeutic concentrations of antimicrobials in health care settings, the presence of subinhibitory antimicrobial residues in the environment and in clinics selects for antimicrobial resistance (AMR), but the underlying genetic repertoire is less well understood. Here, we used machine learning to predict the population doubling time and cell growth yield of 1,407 genetically diverse E. coli strains expanding under exposure to three subinhibitory concentrations of six classes of antimicrobials from single-nucleotide genetic variants, accessory gene variation, and the presence of known AMR genes. We predicted cell growth yields in the held-out test data with an average correlation (Spearman's ρ) of 0.63 (0.36 to 0.81 across concentrations) and cell doubling times with an average correlation of 0.59 (0.32 to 0.92 across concentrations), with moderate increases in sample size unlikely to improve predictions further. This finding points to the remaining missing heritability of growth under antimicrobial exposure being explained by effects that are too rare or weak to be captured unless sample size is dramatically increased, or by effects other than those conferred by the presence of individual single-nucleotide polymorphisms (SNPs) and genes. Predictions based on whole-genome information were generally superior to those based only on known AMR genes and were accurate for AMR resistance at therapeutic concentrations. We pinpointed genes and SNPs determining the predicted growth and thereby recapitulated many known AMR determinants. Finally, we estimated the effect sizes of resistance genes across the entire collection of strains, disclosing the growth effects for known resistance genes in each individual strain. Our results underscore the potential of predictive modeling of growth patterns from genomic data under subinhibitory concentrations of antimicrobials, although the remaining missing heritability poses a challenge for achieving the accuracy and precision required for clinical use. IMPORTANCE Predicting bacterial growth from genome sequences is important for a rapid characterization of strains in clinical diagnostics and to disclose candidate novel targets for anti-infective drugs. Previous studies have dissected the relationship between bacterial growth and genotype in mutant libraries for laboratory strains, yet no study so far has examined the predictive power of genome sequence in natural strains. In this study, we used a high-throughput phenotypic assay to measure the growth of a systematic collection of natural Escherichia coli strains and then employed machine learning models to predict bacterial growth from genomic data under nontherapeutic subinhibitory concentrations of antimicrobials that are common in nonclinical settings. We found a moderate to strong correlation between predicted and actual values for the different collected data sets. Moreover, we observed that the known resistance genes are still effective at sublethal concentrations, pointing to clinical implications of these concentrations.
|
|
| 3. |
- Heyckendorf, Jan, et al.
(författare)
-
What Is Resistance? Impact of Phenotypic versus Molecular Drug Resistance Testing on Therapy for Multi-and Extensively Drug-Resistant Tuberculosis
- 2018
-
Ingår i: Antimicrobial Agents and Chemotherapy. - : AMER SOC MICROBIOLOGY. - 0066-4804 .- 1098-6596. ; 62:2
-
Tidskriftsartikel (refereegranskat)abstract
- Rapid and accurate drug susceptibility testing (DST) is essential for the treatment of multi-and extensively drug-resistant tuberculosis (M/XDR-TB). We compared the utility of genotypic DST assays with phenotypic DST (pDST) using Bactec 960 MGIT or Lowenstein-Jensen to construct M/XDR-TB treatment regimens for a cohort of 25 consecutive M/XDR-TB patients and 15 possible anti-TB drugs. Genotypic DST results from Cepheid GeneXpert MTB/RIF (Xpert) and line probe assays (LPAs; Hain GenoType MTBDRplus 2.0 and MTBDRsl 2.0) and whole-genome sequencing (WGS) were translated into individual algorithmderived treatment regimens for each patient. We further analyzed if discrepancies between the various methods were due to flaws in the genotypic or phenotypic test using MIC results. Compared with pDST, the average agreement in the number of drugs prescribed in genotypic regimens ranged from just 49% (95% confidence interval [ CI], 39 to 59%) for Xpert and 63% (95% CI, 56 to 70%) for LPAs to 93% (95% CI, 88 to 98%) for WGS. Only the WGS regimens did not contain any drugs to which pDST showed resistance. Importantly, MIC testing revealed that pDST likely underestimated the true rate of resistance for key drugs (rifampin, levofloxacin, moxifloxacin, and kanamycin) because critical concentrations (CCs) were too high. WGS can be used to rule in resistance even in M/XDR strains with complex resistance patterns, but pDST for some drugs is still needed to confirm susceptibility and construct the final regimens. Some CCs for pDST need to be reexamined to avoid systematic false-susceptible results in low-level resistant isolates.
|
|
| 4. |
|
|
| 5. |
- Merker, Matthias, et al.
(författare)
-
Phylogenetically informative mutations in genes implicated in antibiotic resistance in Mycobacterium tuberculosis complex
- 2020
-
Ingår i: Genome Medicine. - : BMC. - 1756-994X. ; 12:1
-
Tidskriftsartikel (refereegranskat)abstract
- Background A comprehensive understanding of the pre-existing genetic variation in genes associated with antibiotic resistance in the Mycobacterium tuberculosis complex (MTBC) is needed to accurately interpret whole-genome sequencing data for genotypic drug susceptibility testing (DST). Methods We investigated mutations in 92 genes implicated in resistance to 21 anti-tuberculosis drugs using the genomes of 405 phylogenetically diverse MTBC strains. The role of phylogenetically informative mutations was assessed by routine phenotypic DST data for the first-line drugs isoniazid, rifampicin, ethambutol, and pyrazinamide from a separate collection of over 7000 clinical strains. Selected mutations/strains were further investigated by minimum inhibitory concentration (MIC) testing. Results Out of 547 phylogenetically informative mutations identified, 138 were classified as not correlating with resistance to first-line drugs. MIC testing did not reveal a discernible impact of a Rv1979c deletion shared by M. africanum lineage 5 strains on resistance to clofazimine. Finally, we found molecular evidence that some MTBC subgroups may be hyper-susceptible to bedaquiline and clofazimine by different loss-of-function mutations affecting a drug efflux pump subunit (MmpL5). Conclusions Our findings underline that the genetic diversity in MTBC has to be studied more systematically to inform the design of clinical trials and to define sound epidemiologic cut-off values (ECOFFs) for new and repurposed anti-tuberculosis drugs. In that regard, our comprehensive variant catalogue provides a solid basis for the interpretation of mutations in genotypic as well as in phenotypic DST assays.
|
|
| 6. |
- Moradigaravand, Danesh, et al.
(författare)
-
Prediction of antibiotic resistance in Escherichia coli from large-scale pan-genome data.
- 2018
-
Ingår i: PLoS computational biology. - : Public Library of Science (PLoS). - 1553-7358. ; 14:12
-
Tidskriftsartikel (refereegranskat)abstract
- The emergence of microbial antibiotic resistance is a global health threat. In clinical settings, the key to controlling spread of resistant strains is accurate and rapid detection. As traditional culture-based methods are time consuming, genetic approaches have recently been developed for this task. The detection of antibiotic resistance is typically made by measuring a few known determinants previously identified from genome sequencing, and thus requires the prior knowledge of its biological mechanisms. To overcome this limitation, we employed machine learning models to predict resistance to 11 compounds across four classes of antibiotics from existing and novel whole genome sequences of 1936 E. coli strains. We considered a range of methods, and examined population structure, isolation year, gene content, and polymorphism information as predictors. Gradient boosted decision trees consistently outperformed alternative models with an average accuracy of 0.91 on held-out data (range 0.81-0.97). While the best models most frequently employed gene content, an average accuracy score of 0.90 could be obtained using population structure information alone. Single nucleotide variation data were less useful, and significantly improved prediction only for two antibiotics, including ciprofloxacin. These results demonstrate that antibiotic resistance in E. coli can be accurately predicted from whole genome sequences without a priori knowledge of mechanisms, and that both genomic and epidemiological data can be informative. This paves way to integrating machine learning approaches into diagnostic tools in the clinic.
|
|
| 7. |
- Nakatani, Yoshio, et al.
(författare)
-
Role of Alanine Racemase Mutations in Mycobacterium tuberculosis D-Cycloserine Resistance
- 2017
-
Ingår i: Antimicrobial Agents and Chemotherapy. - : AMER SOC MICROBIOLOGY. - 0066-4804 .- 1098-6596. ; 61:12
-
Tidskriftsartikel (refereegranskat)abstract
- A screening of more than 1,500 drug-resistant strains of Mycobacterium tuberculosis revealed evolutionary patterns characteristic of positive selection for three alanine racemase (Alr) mutations. We investigated these mutations using molecular modeling, in vitro MIC testing, as well as direct measurements of enzymatic activity, which demonstrated that these mutations likely confer resistance to D-cycloserine.
|
|
| 8. |
- Yu, Xia, et al.
(författare)
-
Wild-Type and Non-Wild-Type Mycobacterium tuberculosis MIC Distributions for the Novel Fluoroquinolone Antofloxacin Compared with Those for Ofloxacin, Levofloxacin, and Moxifloxacin
- 2016
-
Ingår i: Antimicrobial Agents and Chemotherapy. - : AMER SOC MICROBIOLOGY. - 0066-4804 .- 1098-6596. ; 60:9, s. 5232-5237
-
Tidskriftsartikel (refereegranskat)abstract
- Antofloxacin (AFX) is a novel fluoroquinolone that has been approved in China for the treatment of infections caused by a variety of bacterial species. We investigated whether it could be repurposed for the treatment of tuberculosis by studying its in vitro activity. We determined the wild-type and non-wild-type MIC ranges for AFX as well as ofloxacin (OFX), levofloxacin (LFX), and moxifloxacin (MFX), using the microplate alamarBlue assay, of 126 clinical Mycobacterium tuberculosis strains from Beijing, China, of which 48 were OFX resistant on the basis of drug susceptibility testing on Lowenstein-Jensen medium. The MIC distributions were correlated with mutations in the quinolone resistance-determining regions of gyrA (Rv0006) and gyrB (Rv0005). Pharmacokinetic/pharmacodynamic (PK/PD) data for AFX were retrieved from the literature. AFX showed lower MIC levels than OFX but higher MIC levels than LFX and MFX on the basis of the tentative epidemiological cutoff values (ECOFFs) determined in this study. All strains with non-wild-type MICs for AFX harbored known resistance mutations that also resulted in non-wild-type MICs for LFX and MFX. Moreover, our data suggested that the current critical concentration of OFX for Lowenstein-Jensen medium that was recently revised by the World Health Organization might be too high, resulting in the misclassification of phenotypically non-wildtype strains with known resistance mutations as wild type. On the basis of our exploratory PK/PD calculations, the current dose of AFX is unlikely to be optimal for the treatment of tuberculosis, but higher doses could be effective.
|
|
