This paper presents a cloud-native architecture for clinical metagenomic diagnostics developed as part of the Horizon Europe project GenDAI. The architecture integrates automated, reproducible, and auditable workflows with deterministic elasticity, compliant data stewardship, explainable Artificial Intelligence (AI), and verifiable reporting. Requirements derived from clinical practice inform a unified modeling, implementation, and evaluation strategy for a modular platform that combines workflow orchestration, data governance, AI-powered modeling, and cloud-native reporting. The system embeds provenance-bydesign, policy-as-code enforcement, and deterministic elasticity across all layers to enable reproducible, compliant, and trustworthy metagenomic diagnostics. This work provides a principled pathway for translating research-grade tools into regulator-ready diagnostic services while maintaining transparency, reproducibility, and long-term trust.
The GenDAI Cloud-Native Infrastructure and Data Stewardship for Clinical Metagenomic Diagnostics / Tamla, Philippe; Krause, Thomas; Hemmje, Matthias; Monti, Flavia; De Luzi, Francesca; Mecella, Massimo; Andrade, Bruno; Buono, Paolo; Molinari, Andrea. - (2025), pp. 7229-7236. ( 2025 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2025 Wuhan; China ) [10.1109/bibm66473.2025.11356811].
The GenDAI Cloud-Native Infrastructure and Data Stewardship for Clinical Metagenomic Diagnostics
Monti, Flavia;De Luzi, Francesca;Mecella, Massimo;
2025
Abstract
This paper presents a cloud-native architecture for clinical metagenomic diagnostics developed as part of the Horizon Europe project GenDAI. The architecture integrates automated, reproducible, and auditable workflows with deterministic elasticity, compliant data stewardship, explainable Artificial Intelligence (AI), and verifiable reporting. Requirements derived from clinical practice inform a unified modeling, implementation, and evaluation strategy for a modular platform that combines workflow orchestration, data governance, AI-powered modeling, and cloud-native reporting. The system embeds provenance-bydesign, policy-as-code enforcement, and deterministic elasticity across all layers to enable reproducible, compliant, and trustworthy metagenomic diagnostics. This work provides a principled pathway for translating research-grade tools into regulator-ready diagnostic services while maintaining transparency, reproducibility, and long-term trust.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


