Federated Microbiome AI for human health – FeMAI

The digital revolution, in particular big data and artificial intelligence (AI), offer new opportunities to transform healthcare. Big data analytics has the potential to combine molecular patient data with electronic health records to fulfil the promises of precision medicine. Systemic marker panels will identify personalized disease mechanisms to treat a disease efficiently. Big data and AI are inevitable technologies to find the needle in the haystack for identifying patients at risk of developing diseases. With FeMAI we want to establish gut microbiome-based machine learning models for the prediction of human health in terms of diagnostic and prognostic. The fundament for big data, however, is data access, which at such dimension raises valid ethical concerns regarding the privacy protection of patient data. They are expressed legally in policies like the General Data Protection Regulation (GDPR), which downscales data sizes from millions of patient data sets to only a few hundred available for AI model learning - reducing their predictive power. With FeMAI we want to address these obstacles by creating a totally privacy preserving federated database network containing decentralized microbiome and clinical data, processed by machine learning tools for human health prediction. The goal is to address the urgently unmet need to access big data in order to pave the road to successful long term collaboration in human health analysis across France, Germany and the EU. Thus FeMAI consortium consists of worldwide recognized experts from AI microbiome development to federated learning.