stackPredAMR publication online

The first results of our project are published in the Journal Bioinformatics Advances. In stackPredAMR - A stacked random forest approach improves AMR phenotype prediction for multiple species and antimicrobial agents the authors present a machine learning framework for the prediction of multiple antimicrobial agents in three different species. The tool was benchmarked on 2,500 publicly available datasets and showed a strong performance. 

The full article can be found online: stackPredAMR – A stacked random forest approach improves AMR phenotype prediction for multiple species and antimicrobial agents | Bioinformatics Advances | Oxford Academic

WIN-KID meets miGenomeSurv

For the second time, members of the WIN-KID consortium joined the yearly miGenomeSurv network meeting. Together they discussed recent advances and future options to improve genomic surveillance and predicition of antimicrobial resistance. Participants of both groups enjoyed the exchange and new perspectives of this growing sctientific network.

First WIN-KID results presented at ESCMID global 2026

Members of the WIN-KID consortium had the opportunity to present their work at the ESCMID Global conference, which took place in Munich in April 2026. In a dedicated E-Poster Flash Session on AI-powered pathogen detection and AMR prediction, PhD student Julian Welling presented stackPred - a stacked random forest approach, which improves AMR phenotype prediction. 

Kick-off Treffen zum Projektstart

©UKM/Fotozentrale/Michael Ibrahim

Zum offiziellen Projektstart sind die drei Partner (Universität Münster, Universität Essen & Ridom GmbH) im Oktober 2024 in Münster zusammengekommen, um sich bei Kaffee und Brötchen kennenzulernen und die ersten Schritte im Projekt zu planen. Nach einer kurzen Vorstellungsrunde und organisatorischen Informationen seitens des Projektträgers, wurden die ersten Aufgaben verteilt und Vorgehensweisen diskutiert.