fedRBE v2

This release enhances fedRBE’s reproducibility by introducing updated workflows for real datasets, including multiomics, downstream analysis tasks such as classification and clustering, and figure generation, while improving setup reliability via a documented Conda/R environment. It also includes small robustness fixes for multi-batch/reference-batch handling, as well as deterministic covariate ordering in the FeatureCloud app.

fedRBE

This releases adds further evaluation of the fedRBE algorithm, while the algorithm itself stays the same.
The fedRBE corrected data was evaluated in supervised (random forest classification) and unsupervised (k-means) learning and the results as well as scripts to reproduce are added to the code in this release.