Background:

Mobile genetic elements (MGEs) play a key role in driving phenotypic innovation and facilitating the acquisition of antimicrobial resistance (AMR). MGEs, especially those derived from commensal bacteria, act as reservoirs of resistance and virulence traits that directly impact human health. For example, pathogenicity islands are MGEs whose integration into bacterial genomes confers virulent properties. The escalating threat of AMR underscores the urgent need to assess emerging mechanisms and develop effective interventions to mitigate their emergence. Predicting MGE integration sites and their associated biological functions remains a formidable challenge due to the diversity of resistance mechanisms and functional roles they encode. Traditional analyses often provide a static view of antimicrobial resistance genes (ARGs), neglecting the dynamic genetic backbones provided by MGEs. This limitation highlights the need for advanced approaches that capture the complexity and dynamics of MGE-mediated resistance. This PhD project proposes to use machine learning strategies to develop surveillance tools that go beyond binary presence/absence detection of ARGs.

Aim/s:

This project aims to improve predictions of AMR risks and inform mitigation strategies by uncovering how MGEs contribute to ARG dissemination within bacterial communities by the development of an XAI framework to elucidate the relationships between MGE properties and their functional impacts, thereby improving interpretability for researchers and policy makers.

AI methods:

An Explainable AI (XAI) framework should be developed to elucidate the relationships between MGE properties, their functional impacts and their organisms they could integrate into.

Keywords:

Mobile genetic elements, antimicrobial resistance, Explainable AI (XAI)