Flimma: a federated and privacy-aware tool for differential gene expression analysis

Genome Biology

Reihaneh Torkzadehmahani¹  Reza Nasirigerdeh²  Daniel Rückert³  Georgios Kaissis⁴  Paolo Tieri⁵  Markus List⁶  Olga Zolotareva⁷  Amir Abbasinejad⁸  David B. Blumenthal⁹  Tobias Frisch^1,10  Nina K. Wenke^1,11  Julian Matschinske^1,11  Mohammad Bakhtiari^1,11  Julian Späth^1,11  Jan Baumbach^1,12 

[1] AI in Medicine and Healthcare, Technical University of Munich, Munich, Germany;AI in Medicine and Healthcare, Technical University of Munich, Munich, Germany;Klinikum rechts der Isar, Technical University of Munich, Munich, Germany;AI in Medicine and Healthcare, Technical University of Munich, Munich, Germany;Klinikum rechts der Isar, Technical University of Munich, Munich, Germany;Biomedical Image Analysis Group, Imperial College London, London, UK;AI in Medicine and Healthcare, Technical University of Munich, Munich, Germany;Klinikum rechts der Isar, Technical University of Munich, Munich, Germany;Biomedical Image Analysis Group, Imperial College London, London, UK;OpenMined, Oxford, UK;CNR National Research Council, IAC Institute for Applied Computing, Rome, Italy;Sapienza University of Rome, Rome, Italy;Chair of Experimental Bioinformatics, TUM School of Life Sciences, Technical University of Munich, Freising, Germany;Chair of Experimental Bioinformatics, TUM School of Life Sciences, Technical University of Munich, Freising, Germany;Institute for Computational Systems Biology, University of Hamburg, Hamburg, Germany;Chair of Experimental Bioinformatics, TUM School of Life Sciences, Technical University of Munich, Freising, Germany;Sapienza University of Rome, Rome, Italy;Department Artificial Intelligence in Biomedical Engineering, Friedrich-Alexander University Erlangen-Nürnberg, Erlangen, Germany;Department of Mathematics and Computer Science, University of Southern Denmark, Odense, Denmark;Institute for Computational Systems Biology, University of Hamburg, Hamburg, Germany;Institute for Computational Systems Biology, University of Hamburg, Hamburg, Germany;Department of Mathematics and Computer Science, University of Southern Denmark, Odense, Denmark;

关键词: Differential expression analysis;    Federated learning;    Privacy of biomedical data;    Meta-analysis;   

DOI  :  10.1186/s13059-021-02553-2

来源: Springer

【 摘 要 】

Aggregating transcriptomics data across hospitals can increase sensitivity and robustness of differential expression analyses, yielding deeper clinical insights. As data exchange is often restricted by privacy legislation, meta-analyses are frequently employed to pool local results. However, the accuracy might drop if class labels are inhomogeneously distributed among cohorts. Flimma (https://exbio.wzw.tum.de/flimma/) addresses this issue by implementing the state-of-the-art workflow limma voom in a federated manner, i.e., patient data never leaves its source site. Flimma results are identical to those generated by limma voom on aggregated datasets even in imbalanced scenarios where meta-analysis approaches fail.

【 授权许可】

CC BY