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.

Isolation and identification of mycobacteria from livestock specimens and milk obtained in Brazil

The prevalence of Mycobacterium bovis and other mycobacterial species in livestock specimens and milk was evaluated. An emphasis was placed upon the distribution of these organisms in milk that is readily available to the public that was either untreated, pasteurized, or treated using ultra high temperature. Twenty-two pathologic specimens from livestock (bovine, swine and bubaline) in five Brazilian states and 128 bovine milk samples from retail markets in the State of São Paulo were examined for mycobacteria. Identification was made by classical biochemical tests, thin layer chromatography of mycolic acids and polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analysis. Mycobacteria were isolated from 15 (68.2 percent) caseous lesions and from 23 (18 percent) milk samples. Eleven isolates were identified as M. bovis, and the remaining 27 nontuberculous mycobacterial isolates were represented by five species and six unidentified rapidly growing mycobacterial strains. The data demonstrate that animal products in Brazil are frequent reservoirs of mycobacteria and may pose a risk to the public