Hazard-based tool for comparative ranking of bacterial isolates according to their antimicrobial resistance (AMR)-associated risk

Le résumé fourni par la source

This Excel-based model (workflow) is a hazard-based tool for comparative ranking of isolates according to their antimicrobial resistance (AMR)-associated risk. Two risk scores are defined: environmental AMR-associated risk (EAAR) and public health AMR-associated risk (HHAAR). The model has been developed for two AMR pathogens: extended-spectrum beta-lactamase-producing (ESBL) Escherichia coli and Salmonella Infantis. Separate worksheets are provided for each risk score and each pathogen.Cells shaded in gray are automatically calculated by the model, whereas cells without a fill color require user input. The required input is an integer representing the number of genes or contigs. In unweighted risk scoring, all risk components contribute equally to the final risk score. In contrast, in weighted risk scoring, user-defined weighting factors can be applied to assign different weights to individual risk components. The model uses a min-max normalization method, in which all dimensions are first transformed to a [0,1] scale. Consequently, the resulting risk scores are inherently dependent on the dataset's extreme values. The final Euclidean distance is normalized by the number of dimensions to ensure that the final risk score also ranges from 0 to 1. The final risk score is unitless. The model is designed to be agnostic to the bacterial species analyzed and the bioinformatic tool/pipeline used. It focuses specifically on hazard characterization based on WGS-derived isolate characteristics and therefore does not constitute a complete risk assessment framework. For example, it does not incorporate disease severity or exposure assessment; rather, it is intended as a hazard ranking tool. To assess the impact of extreme values on isolate ranking, calculations were performed on two isolate datasets. The original dataset comprises only the isolates sequenced in this study. The extended dataset comprises the original dataset supplemented with 25 publicly available genomes for each bacterial species.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.