Predicting antimicrobial resistance in E. coli with discriminative position fused deep learning classifier

Unlocking antimicrobial resistance with multiomics and machine learning

The global antimicrobial resistance (AMR) emergency is driven by complex and evolving molecular mechanisms. Cutting-edge machine learning methods and multiomics technologies can help to combat this crisis by predicting novel AMR biomarkers and outcomes with unprecedented precision and speed, offering critical insights into the molecular underpinnings of AMR.

Relation between serum magnesium and outcome in patients with Escherichia coli sepsis

OBJECTIVE

Escherichia coli (E.coli) is the leading pathogen for deaths associated with antimicrobial resistance, making it the most problematic bacteria for human infections. This study aimed to investigate the association between serum magnesium levels and clinical outcomes in patients with E.coli sepsis.

METHOD

Data of E.coli septic patients were collected from the MIMIC-IV database. Patients were divided into three groups based on tertiles of serum magnesium levels. Three models were utilized, including the raw model (unadjusted), Model I (adjusted for age and gender), and Model II (adjusted for all potential confounding factors). Linear model and two-segment nonlinear model were established to examine the relationship between serum magnesium and 30-day, 60-day, and 90-day mortality rates. Kaplan-Meier survival curve analysis was performed to assess cumulative hazard of mortalities at 30-day, 60-day, 90-day based on tertiles of serum magnesium levels.

RESULTS

A total of 421 E.coli septic patients were included and classified into tertiles: Q1(< 1.6 mg/dL), Q2 (1.6-1.9 mg/dL), Q3(> 1.9 mg/dL). In the Model adjusting for all potential confounders, for every 1 mg/dL increase in serum magnesium, there was a significant increase in 30-day, 60-day, and 90-day mortality rates, with odds ratios of 4.01 (95% CI 1.22-13.19, P = 0.022), 4.81 (95% CI 1.59-14.53, P = 0.005), and 4.45 (95% CI 1.52-12.96, P = 0.006) respectively. And linear model is more suitable for describing the relationship between serum magnesium levels and clinical outcomes. Kaplan-Meier analysis revealed that the cumulative hazard of mortalities at 30-day, 60-day, 90-day increased with the prolongation of hospital stay, particularly in the group with the highest serum magnesium level.

CONCLUSION

Increased level of serum magnesium is significantly associated with increased risk of 30-day, 60-day and 90-day mortality in a population of septic patients with E.coli infection.

Ecological prevalence and genomic characterization of Salmonella isolated from selected poultry farms in Jiangxi province, China

Non-typhoidal Salmonella (NTS), particularly antimicrobial-resistant serovars, remains the major source of foodborne bacterial illnesses. Raw chicken is the leading cause of human salmonellosis. In this study, we evaluated the prevalence, antimicrobial resistance profiles, and genomic features of 143/1,800 (7.94%) Salmonella strains isolated from poultry farms in five major regions of Jiangxi province, China, between 2022 and 2023 using Whole genome sequencing (WGS). Among Salmonella isolates, the most common serovars were Infantis (ST32) and Enteritidis (ST11). Resistance to amoxicillin and tetracycline was the most prevalent, with 60.84% of Salmonella isolates exhibiting a multi-drug resistance (MDR) pattern. The detection of antimicrobial-resistant genes (ARGs) examined was aligned with the resistant phenotypes found. A total of 61 ARGs were identified, with aph(3')-Ia, qnrS1, aph(3'')-Ib, and tetA being the prominent ARGs. Furthermore, 24 beta-lactam genes were also identified, including blaTEM, blaSHV, and blaCTX-M. The number of ARGs and the distribution of serovars varied according to the year, farms, and cities. Salmonella isolates carried 13 heavy metal resistance genes (HMRGs) and two biocide resistance genes, with pcoS being the most prevalent. A total of 145 virulence genes and 19 plasmids were found, with serovars Infantis and Enteritidis having the most virulence genes. The high occurrence of MDR Salmonella in this study, particularly carrying numerous mobile genetic elements (MGEs), posed a serious threat to food safety and public health, emphasizing the need to improve poultry farm hygiene to decrease contamination and transmission.

Clinical microbiology and artificial intelligence: Different applications, challenges, and future prospects

Conventional clinical microbiological techniques are enhanced by the introduction of artificial intelligence (AI). Comprehensive data processing and analysis enabled the development of curated datasets that has been effectively used in training different AI algorithms. Recently, a number of machine learning (ML) and deep learning (DL) algorithms are developed and evaluated using diverse microbiological datasets. These datasets included spectral analysis (Raman and MALDI-TOF spectroscopy), microscopic images (Gram and acid fast stains), and genomic and protein sequences (whole genome sequencing (WGS) and protein data banks (PDBs)). The primary objective of these algorithms is to minimize the time, effort, and expenses linked to conventional analytical methods. Furthermore, AI algorithms are incorporated with quantitative structure-activity relationship (QSAR) models to predict novel antimicrobial agents that address the continuing surge of antimicrobial resistance. During the COVID-19 pandemic, AI algorithms played a crucial role in vaccine developments and the discovery of new antiviral agents, and introduced potential drug candidates via drug repurposing. However, despite their significant benefits, the implementation of AI encounters various challenges, including ethical considerations, the potential for bias, and errors related to data training. This review seeks to provide an overview of the most recent applications of artificial intelligence in clinical microbiology, with the intention of educating a wider audience of clinical practitioners regarding the current uses of machine learning algorithms and encouraging their implementation. Furthermore, it will discuss the challenges related to the incorporation of AI into clinical microbiology laboratories and examine future opportunities for AI within the realm of infectious disease epidemiology.

Predicting total healthcare demand using machine learning: separate and combined analysis of predisposing, enabling, and need factors

OBJECTIVE

Predicting healthcare demand is essential for effective resource allocation and planning. This study applies Andersen's Behavioral Model of Health Services Use, focusing on predisposing, enabling, and need factors, using data from the 2022 Turkey Health Survey by TUIK. Machine learning methods provide a powerful approach to analyze these factors and their combined impact on healthcare utilization, offering valuable insights for health policy.

METHODS

Seven different machine learning models-Decision Tree, Random Forest, Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Logistic Regression, XGBoost, and Gradient Boosting-were utilized. Feature selection was conducted to identify the most significant factors influencing healthcare demand. The models were evaluated for accuracy and generalization ability using performance metrics such as recall, precision, F1 score, and ROC AUC.

RESULTS

The study identified key features affecting healthcare demand. For predisposing factors, gender, educational level, and age group were significant. Enabling factors included treatment costs, community interest, and payment difficulties. Need factors were influenced by smoking status, chronic diseases, and overall health status. The models demonstrated high recall (approximately 0.90) and strong F1 scores (ranging from 0.87 to 0.88), indicating a balanced performance between precision and recall. Among the models, Gradient Boosting, XGBoost, and Logistic Regression consistently outperformed others, achieving the highest predictive accuracy. Random Forest and SVM also performed well, showing robust classification capability.

CONCLUSIONS

The findings highlight the effectiveness of machine learning methods in predicting healthcare demand, providing valuable insights for health policy and resource allocation. Gradient Boosting, XGBoost, and Logistic Regression emerged as the most reliable models, demonstrating superior generalization and classification performance. Understanding the separate and combined effects of predisposing, enabling, and need factors on healthcare demand can contribute to more efficient and data-driven healthcare planning, facilitating strategic decision-making in resource allocation and service delivery.

Avian-specific Salmonella transition to endemicity is accompanied by localized resistome and mobilome interaction

Bacterial regional demonstration after global dissemination is an essential pathway for selecting distinct finesses. However, the evolution of the resistome during the transition to endemicity remains unaddressed. Using the most comprehensive whole-genome sequencing dataset of Salmonella enterica serovar Gallinarum (S. Gallinarum) collected from 15 countries, including 45 newly recovered samples from two related local regions, we established the relationship among avian-specific pathogen genetic profiles and localization patterns. Initially, we revealed the international transmission and evolutionary history of S. Gallinarum to recent endemicity through phylogenetic analysis conducted using a spatiotemporal Bayesian framework. Our findings indicate that the independent acquisition of the resistome via the mobilome, primarily through plasmids and transposons, shapes a unique antimicrobial resistance profile among different lineages. Notably, the mobilome-resistome combination among distinct lineages exhibits a geographical-specific manner, further supporting a localized endemic mobilome-driven process. Collectively, this study elucidates resistome adaptation in the endemic transition of an avian-specific pathogen, likely driven by the localized farming style, and provides valuable insights for targeted interventions.

Prediction of inhibitory peptides against E.coli with desired MIC value

In the past, several methods have been developed for predicting antibacterial and antimicrobial peptides, but only limited attempts have been made to predict their minimum inhibitory concentration (MIC) values. In this study, we developed predictive models for MIC values of antibacterial peptides against Escherichia coli (E. coli), comprised of 3143 peptides for training and 786 peptides for validation, with experimentally determined MIC values. We found that the Composition Enhanced Transition and Distribution (CeTD) attributes significantly correlate with MIC values. Initially, we attempted to estimate MIC using BLAST similarity searches but found them inadequate. Subsequently, we employed machine learning regression models that integrated various features, including peptide composition, binary profiles and embeddings from large language models. Feature selection techniques, particularly mRMR, were utilized to refine our model inputs. Our Random Forest regressor built using default parameters achieved a correlation coefficient (R) of 0.78, R2 of 0.59, and RMSE of 0.53 on the validation set. Our best model outperformed existing methods when benchmarked on an independent dataset of 498 anti-E. coli peptides. Additionally, we screened anti-E. coli proteins in the proteomes of three probiotic bacterial strains and created a web-based platform, "EIPpred", enabling users to design peptides with desired MIC values ( https://webs.iiitd.edu.in/raghava/eippred ).

A multi-channel electrochemical biosensor based on polyadenine tetrahedra for the detection of multiple drug resistance genes

Antimicrobial resistance poses a serious threat to human health due to the high morbidity and mortality caused by drug-resistant microbial infections. Therefore, the development of rapid, sensitive and selective identification methods is key to improving the survival rate of patients. In this paper, a sandwich-type electrochemical DNA biosensor based on a polyadenine-DNA tetrahedron probe was constructed. The key experimental conditions were optimized, including the length of polyadenine, the concentration of the polyadenine DNA tetrahedron, the concentration of the signal probe and the hybridization time. At the same time, poly-avidin-HRP80 was used to enhance the electrochemical detection signal. Finally, excellent biosensor performance was achieved, and the detection limit for the synthetic DNA target was as low as 1 fM. In addition, we verified the practicability of the system by analyzing E. coli with the MCR-1 plasmid and realized multi-channel detection of the drug resistance genes MCR-1, blaNDM, blaKPC and blaOXA. With the ideal electrochemical interface, the polyA-based biosensor exhibits excellent stability, which provides powerful technical support for the rapid detection of antibiotic-resistant strains in the field.

Tackling the Antimicrobial Resistance "Pandemic" with Machine Learning Tools: A Summary of Available Evidence

Antimicrobial resistance is recognised as one of the top threats healthcare is bound to face in the future. There have been various attempts to preserve the efficacy of existing antimicrobials, develop new and efficient antimicrobials, manage infections with multi-drug resistant strains, and improve patient outcomes, resulting in a growing mass of routinely available data, including electronic health records and microbiological information that can be employed to develop individualised antimicrobial stewardship. Machine learning methods have been developed to predict antimicrobial resistance from whole-genome sequencing data, forecast medication susceptibility, recognise epidemic patterns for surveillance purposes, or propose new antibacterial treatments and accelerate scientific discovery. Unfortunately, there is an evident gap between the number of machine learning applications in science and the effective implementation of these systems. This narrative review highlights some of the outstanding opportunities that machine learning offers when applied in research related to antimicrobial resistance. In the future, machine learning tools may prove to be superbugs' kryptonite. This review aims to provide an overview of available publications to aid researchers that are looking to expand their work with new approaches and to acquaint them with the current application of machine learning techniques in this field.

Salmonellosis outbreak archive in China: data collection and assembly

Infectious disease outbreaks transcend the medical and public health realms, triggering widespread panic and impeding socio-economic development. Considering that self-limiting diarrhoea of sporadic cases is usually underreported, the Salmonella outbreak (SO) study offers a unique opportunity for source tracing, spatiotemporal correlation, and outbreak prediction. To summarize the pattern of SO and estimate observational epidemiological indicators, 1,134 qualitative reports screened from 1949 to 2023 were included in the systematic review dataset, which contained a 506-study meta-analysis dataset. In addition to the dataset comprising over 50 columns with a total of 46,494 entries eligible for inclusion in systematic reviews or input into prediction models, we also provide initial literature collection datasets and datasets containing socio-economic and climate information for relevant regions. This study has a broad impact on advancing knowledge regarding epidemic trends and prevention priorities in diverse salmonellosis outbreaks and guiding rational policy-making or predictive modeling to mitigate the infringement upon the right to life imposed by significant epidemics.