Burankova Yuliya, Abele Miriam, Bakhtiari Mohammad, von Toerne Christine, Barth Teresa K, Schweizer Lisa, Giesbertz Pieter, Schmidt Johannes R, Kalkhof Stefan, Müller-Deile Janina, van Veelen Peter A, Mohammed Yassene, Hammer Elke, Arend Lis, Adamowicz Klaudia, Laske Tanja, Hartebrodt Anne, Frisch Tobias, Meng Chen, Matschinske Julian, Späth Julian, Röttger Richard, Schwämmle Veit, Hauck Stefanie M, Lichtenthaler Stefan F, Imhof Axel, Mann Matthias, Ludwig Christina, Kuster Bernhard, Baumbach Jan, Zolotareva Olga
Chair of Proteomics and Bioanalytics, TUM School of Life Sciences, Technical University of Munich, Freising, Germany.
Institute for Computational Systems Biology, University of Hamburg, Hamburg, Germany.
Nat Comput Sci. 2025 Aug;5(8):675-688. doi: 10.1038/s43588-025-00832-7. Epub 2025 Jul 11.
Quantitative mass spectrometry has revolutionized proteomics by enabling simultaneous quantification of thousands of proteins. Pooling patient-derived data from multiple institutions enhances statistical power but raises serious privacy concerns. Here we introduce FedProt, the first privacy-preserving tool for collaborative differential protein abundance analysis of distributed data, which utilizes federated learning and additive secret sharing. In the absence of a multicenter patient-derived dataset for evaluation, we created two: one at five centers from E. coli experiments and one at three centers from human serum. Evaluations using these datasets confirm that FedProt achieves accuracy equivalent to the DEqMS method applied to pooled data, with completely negligible absolute differences no greater than 4 × 10. By contrast, -logP computed by the most accurate meta-analysis methods diverged from the centralized analysis results by up to 25-26.
定量质谱技术通过能够同时对数千种蛋白质进行定量，彻底改变了蛋白质组学。汇集来自多个机构的患者衍生数据可增强统计能力，但引发了严重的隐私问题。在此，我们介绍FedProt，这是首个用于分布式数据协作差异蛋白质丰度分析的隐私保护工具，它利用联邦学习和加法秘密共享。在缺乏用于评估的多中心患者衍生数据集的情况下，我们创建了两个数据集：一个来自大肠杆菌实验的五个中心，另一个来自人血清的三个中心。使用这些数据集进行的评估证实，FedProt实现的准确度与应用于汇集数据的DEqMS方法相当，绝对差异完全可忽略不计，不超过4×10。相比之下，最准确的荟萃分析方法计算出的-logP与集中分析结果的差异高达25-26。
