PHT-PADME is a distributed infrastructure based on the FAIR principles (Findability, Accessability, Interoperability, Reproducibility) from the GoFAIR initiative.

PADME enables secure and privacy preserving analysis of patient data with federated and incremental learning.

It provides services to researchers, data scientists and AI and Machine Learning specialists to run their models following the distributed analytics paradigm of ‘bringing analysis to data’ instead of the conventional centralized analytics (‘bring data to analysis’).

Moreover, it enables analysis of sensitive data, which otherwise will not be available for analysis. Target research artifacts are data science algorithms and pipelines to run in federated or incremental settings. The repeatability and provenance of execution is ensured by rich metadata.

PADME uses private sensitive data stored in secure environments and dockerized/containerized data analytics algorithms.

The PHT- PADME platforms have to be installed in research data networks. Central services can be run either by NFDI4DS or support can be provided to hosting partners.

Link: /services/padme