Prediction and interpretation of antibiotic-resistance genes occurrence at recreational beaches using machine learning models.

Antibiotic-resistant bacteria and antibiotic resistance genes (ARGs) are pollutants of worldwide concern that seriously threaten public health and ecosystems. Machine learning (ML) prediction models have been applied to predict ARGs in beach waters. However, the existing studies were conducted at a single location and had low prediction performance. Moreover, ML models are "black boxes" that do not reveal their predictions' internal nuances and mechanisms. This lack of transparency and trust can result in serious consequences when using these models in high-stakes decisions. In this study, we developed a gradient boosted regression tree based (GBRT) ML model and then described its behavior using six explainable artificial intelligence (XAI) model-agnostic explanation methods. We used hydro-meteorological and qPCR data from the beaches in South Korea and Pakistan and developed ML prediction models for aac (6'-lb-cr), sul1, and tetX with 10-fold time-blocked cross-validation performances of 4.9, 2.06 and 4.4 root mean squared logarithmic error, respectively. We then analyzed the local and global behavior of the developed ML model using four interpretation methods. The developed ML models showed that water temperature, precipitation and tide are the most important predictors for prediction of ARGs at recreational beaches. We show that the model-agnostic interpretation methods not only explain the behavior of the ML model but also provide insights into the behavior of the ML model under new unseen conditions. Moreover, these post-processing techniques can be a debugging tool for ML-based modeling.

Application of Decision-Tree-Based Machine Learning Algorithms for Prediction of Antimicrobial Resistance.

Timely and efficacious antibiotic treatment depends on precise and quick in silico antimicrobial-resistance predictions. Limited treatment choices due to antimicrobial resistance (AMR) highlight the necessity to optimize the available diagnostics. AMR can be explicitly anticipated on the basis of genome sequence. In this study, we used transcriptomes of 410 multidrug-resistant isolates of Pseudomonas aeruginosa. We trained 10 machine learning (ML) classifiers on the basis of data on gene expression (GEXP) information and generated predictive models for meropenem, ciprofloxacin, and ceftazidime drugs. Among all the used ML models, four models showed high F1-score, accuracy, precision, and specificity compared with the other models. However, RandomForestClassifier showed a moderate F1-score (0.6), precision (0.61), and specificity (0.625) for ciprofloxacin. In the case of ceftazidime, RidgeClassifier performed well and showed F1-score (0.652), precision (0.654), and specificity (0.652) values. For meropenem, KNeighborsClassifier exhibited moderate F1-score (0.629), precision (0.629), and specificity (0.629). Among these three antibiotics, GEXP data on meropenem and ceftazidime improved diagnostic performance. The findings will pave the way for the establishment of a resistance profiling tool that can predict AMR on the basis of transcriptomic markers.

Prediction of antimicrobial minimal inhibitory concentrations for Neisseria gonorrhoeae using machine learning models.

The lowest concentration of an antimicrobial agent that can inhibit the visible growth of a microorganism after overnight incubation is called as minimum inhibitory concentration (MIC) and the drug prescriptions are made on the basis of MIC data to ensure successful treatment outcomes. Therefore, reliable antimicrobial susceptibility data is crucial, and it will help clinicians about which drug to prescribe. Although few prediction studies based on strategies have been conducted, however, no single machine learning (ML) modelling has been carried out to predict MICs in N. gonorrhoeae. In this study, we propose a ML based approach that can predict MICs of a specific antibiotic using unitigs sequences data. We retrieved N. gonorrhoeae genomes from European Nucleotide Archive and NCBI and analysed them combined with their respective MIC data for cefixime, ciprofloxacin, and azithromycin and then we constructed unitigs by using de Brujin graphs. We built and compared 35 different ML regression models to predict MICs. Our results demonstrate that RandomForest and CATBoost models showed best performance in predicting MICs of the three antibiotics. The coefficient of determination, R2, (a statistical measure of how well the regression predictions approximate the real data points) for cefixime, ciprofloxacin, and azithromycin was 0.75787, 0.77241, and 0.79009 respectively using RandomForest. For CATBoost model, the R2 value was 0.74570, 0.77393, and 0.79317 for cefixime, ciprofloxacin, and azithromycin respectively. Lastly, using feature importance, we explore the important genomic regions identified by the models for predicting MICs. The major mutations which are responsible for resistance against these three antibiotics were chosen by ML models as a top feature in case of each antibiotics. CATBoost, DecisionTree, GradientBoosting, and RandomForest regression models chose the same unitigs which are responsible for resistance. This unitigs-based strategy for developing models for MIC prediction, clinical diagnostics, and surveillance can be applicable for other critical bacterial pathogens.

Prevalence of clinical malaria and household characteristics of patients in tribal districts of Pakistan.

BACKGROUND: Malaria, disproportionately affects poor people more than any other disease of public health concern in developing countries. In resource-constrained environments, monitoring the occurrence of malaria is essential for the success of national malaria control programs. Militancy and military conflicts have been a major challenge in monitoring the incidence and controlling malaria and other emerging infectious diseases. The conflicts and instability in Afghanistan have resulted in the migration of refugees into the war-torn tribal districts of Pakistan's Khyber Pakhtunkhwa (KPK) province and the possible introduction of many contagious epidemics. Although malaria is very common in all tribal districts, molecular, clinical and epidemiological data are scarce in these high-burden districts. Therefore, for the proper surveillance, detection, and control of malaria, obtaining and analyzing reliable data in these districts is essential. METHODOLOGY/PRINCIPAL FINDINGS: All 1,127 malaria-suspected patients were sampled within the transmission season in the tribal districts of KPK province between March 2016 to December 2018. After a detailed demographic and clinical investigation of malaria-suspected patients, the data were recorded. The data of the control group was collected simultaneously at the same site. They were considered as uncomplicated cases for statistical analyses. Blood samples were collected from malaria-suspected patients for the detection of Plasmodium species using microscopy and nested PCR (nPCR). Microscopy and nPCR examination detected 78% (n = 882) and 38% (n = 429) Plasmodium-positive patients, respectively. Among1,127 of 429nPCR detected cases with both species of malaria, the frequency of complications was as follows: anemia (n = 71; 16.5%), decompensated shock (n = 40; 9%), hyperpyrexia (n = 117; 27%), hyperparasitaemia (n = 49; 11%) hypoglycemia (n = 45; 10.5%), jaundice (n = 54; 13%), multiple convulsions (n = 37; 9%), and petechia (n = 16; 4%). We observed that 37% (n = 157 out of 429) of those patients infected by both Plasmodium species were children between the ages of 1 and 15 years old. The results revealed that Bajaur (24%), Kurram (20%), and Khyber (18%) districtshada higher proportion of P. vivax than P. falciparum cases. Most of the malaria cases were males (74%). Patients infected by both Plasmodium species tended to less commonly have received formal education and ownership of wealth indicators (e.g., fridge, TV set) was lower. CONCLUSIONS/SIGNIFICANCE: Malaria in tribal districts of the KPK province largely affects young males. P. vivax is a major contributor to the spread of malaria in the area, including severe malaria. We observed a high prevalence of P. vivax in the Bajaur district. Children were the susceptible population to malaria infections whereas they were the least expected to use satisfactory prevention strategies. A higher level of education, a possession of TV sets, the use of bed nets, the use of repellent fluids, and fridges were all associated with protection from malaria. An increased investment in socio-economic development, a strong health infrastructure, and malaria education are key interventions to reduce malaria in the tribal districts.

The novel metallo-ß-lactamase PNGM-1 from a deep-sea sediment metagenome: crystallization and X-ray crystallographic analysis.

Metallo-ß-lactamases (MBLs) are present in major Gram-negative pathogens and environmental species, and pose great health risks because of their ability to hydrolyze the ß-lactam rings of antibiotics such as carbapenems. PNGM-1 was the first reported case of a subclass B3 MBL protein that was identified from a metagenomic library from deep-sea sediments that predate the antibiotic era. In this study, PNGM-1 was overexpressed, purified and crystallized. Crystals of native and selenomethionine-substituted PNGM-1 diffracted to 2.10 and 2.30 Šresolution, respectively. Both the native and the selenomethionine-labelled PNGM-1 crystals belonged to the monoclinic space group P21, with unit-cell parameters a = 122, b = 83, c = 163 Å, ß = 110°. Matthews coefficient (VM) calculations suggested the presence of 6-10 molecules in the asymmetric unit, corresponding to a solvent content of ∼31-58%. Structure determination is currently in progress.

PNGM-1, a novel subclass B3 metallo-ß-lactamase from a deep-sea sediment metagenome.

OBJECTIVES: In order to find antimicrobial resistance gene(s) pre-dating the use of antibiotics through metagenomics, functional screening of a metagenomic library from the deep-seep sediments of Edison Seamount (ca. 10000 years old) was performed. METHODS: Among 60 antimicrobial-resistant clones, a single clone with the highest minimum inhibitory concentration (MIC) for ampicillin was selected. Sequence analysis revealed a new metallo-ß-lactamase (MBL) gene, designated as blaPNGM-1. PNGM-1 retains a zinc ion-binding motif (H116XH118XD120H121, H196 and H263), conserved in subclass B3 MBLs. The catalytic parameters of purified PNGM-1 and the MICs of ß-lactams for Escherichia coli TOP10 transformants harbouring the blaPNGM-1 gene were assessed. RESULTS: Antimicrobial susceptibility testing indicated reduced susceptibility to penicillins, narrow- and extended-spectrum cephalosporins, and carbapenems in E. coli TOP10 transformants harbouring the blaPNGM-1 gene. In addition, kinetic analyses revealed that PNGM-1 hydrolysed almost all ß-lactams. CONCLUSIONS: The PNGM-1 enzyme is the first case of a subclass B3 MBL derived from a functional metagenomic library of a deep-sea sediment that pre-dates the antibiotic era.

Epidemiology and Clinical Burden of Malaria in the War-Torn Area, Orakzai Agency in Pakistan.

BACKGROUND: Military conflict has been a major challenge in the detection and control of emerging infectious diseases such as malaria. It poses issues associated with enhancing emergence and transmission of infectious diseases by destroying infrastructure and collapsing healthcare systems. The Orakzai agency in Pakistan has witnessed a series of intense violence and destruction. Military conflicts and instability in Afghanistan have resulted in the migration of refugees into the area and possible introduction of many infectious disease epidemics. Due to the ongoing violence and Talibanization, it has been a challenge to conduct an epidemiological study. METHODOLOGY/PRINCIPAL FINDINGS: All patients were sampled within the transmission season. After a detailed clinical investigation of patients, data were recorded. Baseline venous blood samples were taken for microscopy and nested polymerase chain reaction (nPCR) analysis. Plasmodium species were detected using nested PCR (nPCR) and amplification of the small subunit ribosomal ribonucleic acid (ssrRNA) genes using the primer pairs. We report a clinical assessment of the epidemic situation of malaria caused by Plasmodium vivax (86.5%) and Plasmodium falciparum (11.79%) infections with analysis of complications in patients such as decompensated shock (41%), anemia (8.98%), hypoglycaemia (7.3%), multiple convulsions (6.7%), hyperpyrexia (6.17%), jaundice (5%), and hyperparasitaemia (4.49%). CONCLUSIONS/SIGNIFICANCE: This overlooked distribution of P. vivax should be considered by malaria control strategy makers in the world and by the Government of Pakistan. In our study, children were the most susceptible population to malaria infection while they were the least expected to use satisfactory prevention strategies in such a war-torn deprived region. Local health authorities should initiate malaria awareness programs in schools and malaria-related education should be further promoted at the local level reaching out to both children and parents.