Dataset Information

Discrimination of contagious and environmental strains of Streptococcus uberis in dairy herds by means of mass spectrometry and machine-learning.

ABSTRACT: Streptococcus uberis is one of the most common pathogens of clinical mastitis in the dairy industry. Knowledge of pathogen transmission route is essential for the selection of the most suitable intervention. Here we show that spectral profiles acquired from clinical isolates using matrix-assisted laser desorption ionization/time of flight (MALDI-TOF) can be used to implement diagnostic classifiers based on machine learning for the successful discrimination of environmental and contagious S. uberis strains. Classifiers dedicated to individual farms achieved up to 97.81% accuracy at cross-validation when using a genetic algorithm, with Cohen's kappa coefficient of 0.94. This indicates the potential of the proposed methodology to successfully support screening at the herd level. A global classifier developed on merged data from 19 farms achieved 95.88% accuracy at cross-validation (kappa 0.93) and 70.67% accuracy at external validation (kappa 0.34), using data from another 10 farms left as holdout. This indicates that more work is needed to develop a screening solution successful at the population level. Significant MALDI-TOF spectral peaks were extracted from the trained classifiers. The peaks were found to correspond to bacteriocin and ribosomal proteins, suggesting that immunity, growth and competition over nutrients may be correlated to the different transmission routes.

SUBMITTER: Esener N

PROVIDER: S-EPMC6269454 | biostudies-literature | 2018 Nov

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Discrimination of contagious and environmental strains of Streptococcus uberis in dairy herds by means of mass spectrometry and machine-learning.

Esener Necati N Green Martin J MJ Emes Richard D RD Jowett Benjamin B Davies Peers L PL Bradley Andrew J AJ Dottorini Tania T

Scientific reports 20181130 1

Streptococcus uberis is one of the most common pathogens of clinical mastitis in the dairy industry. Knowledge of pathogen transmission route is essential for the selection of the most suitable intervention. Here we show that spectral profiles acquired from clinical isolates using matrix-assisted laser desorption ionization/time of flight (MALDI-TOF) can be used to implement diagnostic classifiers based on machine learning for the successful discrimination of environmental and contagious S. uber ...[more]

PMID: 30504894

Similar Datasets

Project description:Streptococcus uberis is considered one of the most important pathogens associated with bovine mastitis. While traditionally acknowledged as an environmental pathogen, S. uberis has been shown to adopt a contagious epidemiological pattern in several dairy herds. Since different control strategies are employed depending on the mode of transmission, in-depth studies of S. uberis populations are essential to determine the best practices to control this pathogen. In this work, we optimized and validated a dot blot platform, combined with automatic image analysis, to rapidly assess the population structure of infective S. uberis, and evaluated its efficiency when compared to multilocus sequence analysis (MLSA) genotyping. Two dairy herds with prevalent S. uberis infections were followed in a 6 month period, in order to collect and characterize isolates from cows with persistent infections. These herds, located in Portugal (Barcelos and Maia regions), had similar management practices, with the herd from Barcelos being smaller and having a better milking parlor management, since infected cow segregation was immediate. A total of 54 S. uberis isolates were obtained from 24 different cows from the two herds. To overcome operator-dependent analysis of the dot blots and increase the technique's consistency and reliability, the hybridization signals were converted into probability values, with average probabilities higher than 0.5 being considered positive results. These data allowed to confirm the isolates' identity as S. uberis using taxa-specific markers and to determine the presence of virulence- and antibiotic resistance-related genes. In addition, MLSA allowed to disclose the most prevalent S. uberis clonal lineages in both herds. Seven different clusters were identified, with Barcelos showing a high clonal diversity and Maia a dominant lineage infecting most cows, suggesting distinct epidemiological patterns, with S. uberis displaying an environmental or contagious transmission pattern depending on the herd. Overall, this work showed the utility of dot blot and MLSA to characterize population structure and epidemiological patterns of mastitis-causing S. uberis. This approach allowed to disclose prevalent virulence patterns and clonal lineages of S. uberis in two distinct herds, and gain insights on the impact of herd management practices on pathogen population structure.

Project description:The objectives of this study are to determine the occurrence of antimicrobial resistance (AMR) genes using whole-genome sequence (WGS) of Streptococcus uberis (S. uberis) and Streptococcus dysgalactiae (S. dysgalactiae) isolates, recovered from dairy cows in the Canadian Maritime Provinces. A secondary objective included the exploration of the association between phenotypic AMR and the genomic characteristics (genome size, guanine-cytosine content, and occurrence of unique gene sequences). Initially, 91 isolates were sequenced, and of these isolates, 89 were assembled. Furthermore, 16 isolates were excluded due to larger than expected genomic sizes (>2.3 bp × 1,000 bp). In the final analysis, 73 were used with complete WGS and minimum inhibitory concentration records, which were part of the previous phenotypic AMR study, representing 18 dairy herds from the Maritime region of Canada (1). A total of 23 unique AMR gene sequences were found in the bacterial genomes, with a mean number of 8.1 (minimum: 5; maximum: 13) per genome. Overall, there were 10 AMR genes [ANT(6), TEM-127, TEM-163, TEM-89, TEM-95, Linb, Lnub, Ermb, Ermc, and TetS] present only in S. uberis genomes and 2 genes unique (EF-TU and TEM-71) to the S. dysgalactiae genomes; 11 AMR genes [APH(3'), TEM-1, TEM-136, TEM-157, TEM-47, TetM, bl2b, gyrA, parE, phoP, and rpoB] were found in both bacterial species. Two-way tabulations showed association between the phenotypic susceptibility to lincosamides and the presence of linB (P = 0.002) and lnuB (P < 0.001) genes and the between the presence of tetM (P = 0.015) and tetS (P = 0.064) genes and phenotypic resistance to tetracyclines only for the S. uberis isolates. The logistic model showed that the odds of resistance (to any of the phenotypically tested antimicrobials) was 4.35 times higher when there were >11 AMR genes present in the genome, compared with <7 AMR genes (P < 0.001). The odds of resistance was lower for S. dysgalactiae than S. uberis (P = 0.031). When the within-herd somatic cell count was >250,000 cells/mL, a trend toward higher odds of resistance compared with the baseline category of <150,000 cells/mL was observed. When the isolate corresponded to a post-mastitis sample, there were lower odds of resistance when compared with non-clinical isolates (P = 0.01). The results of this study showed the strength of associations between phenotypic AMR resistance of both mastitis pathogens and their genotypic resistome and other epidemiological characteristics.

Dataset Information

Discrimination of contagious and environmental strains of Streptococcus uberis in dairy herds by means of mass spectrometry and machine-learning.

Publications

Discrimination of contagious and environmental strains of Streptococcus uberis in dairy herds by means of mass spectrometry and machine-learning.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets