TY - JOUR
T1 - dsMTL
T2 - a computational framework for privacy-preserving, distributed multi-task machine learning
AU - The COMMITMENT Consortium
AU - Cao, Han
AU - Zhang, Youcheng
AU - Baumbach, Jan
AU - Burton, Paul R.
AU - Dwyer, Dominic
AU - Koutsouleris, Nikolaos
AU - Matschinske, Julian
AU - Marcon, Yannick
AU - Rajan, Sivanesan
AU - Rieg, Thilo
AU - Ryser-Welch, Patricia
AU - Späth, Julian
AU - Herrmann, Carl
AU - Schwarz, Emanuel
AU - Alnæs, Dag
AU - Andreassen, Ole A.
AU - Chen, Junfang
AU - Degenhardt, Franziska
AU - Doncevic, Daria
AU - Eils, Roland
AU - Erdmann, Jeanette
AU - Hofmann-Apitius, Martin
AU - Kodamullil, Alpha T.
AU - Khuntia, Adyasha
AU - Mucha, Sören
AU - Nöthen, Markus M.
AU - Paul, Riya
AU - Pedersen, Mads L.
AU - Schunkert, Heribert
AU - Tost, Heike
AU - Westlye, Lars T.
AU - Meyer-Lindenberg, Andreas
N1 - Publisher Copyright:
© The Author(s) 2022. Published by Oxford University Press.
PY - 2022/11/1
Y1 - 2022/11/1
N2 - Motivation: In multi-cohort machine learning studies, it is critical to differentiate between effects that are reproducible across cohorts and those that are cohort-specific. Multi-task learning (MTL) is a machine learning approach that facilitates this differentiation through the simultaneous learning of prediction tasks across cohorts. Since multi-cohort data can often not be combined into a single storage solution, there would be the substantial utility of an MTL application for geographically distributed data sources. Results: Here, we describe the development of ‘dsMTL’, a computational framework for privacy-preserving, distributed multi-task machine learning that includes three supervised and one unsupervised algorithms. First, we derive the theoretical properties of these methods and the relevant machine learning workflows to ensure the validity of the software implementation. Second, we implement dsMTL as a library for the R programming language, building on the DataSHIELD platform that supports the federated analysis of sensitive individual-level data. Third, we demonstrate the applicability of dsMTL for comorbidity modeling in distributed data. We show that comorbidity modeling using dsMTL outperformed conventional, federated machine learning, as well as the aggregation of multiple models built on the distributed datasets individually. The application of dsMTL was computationally efficient and highly scalable when applied to moderate-size (n < 500), real expression data given the actual network latency. Availability and implementation: dsMTL is freely available at https://github.com/transbioZI/dsMTLBase (server-side package) and https://github.com/transbioZI/dsMTLClient (client-side package).
AB - Motivation: In multi-cohort machine learning studies, it is critical to differentiate between effects that are reproducible across cohorts and those that are cohort-specific. Multi-task learning (MTL) is a machine learning approach that facilitates this differentiation through the simultaneous learning of prediction tasks across cohorts. Since multi-cohort data can often not be combined into a single storage solution, there would be the substantial utility of an MTL application for geographically distributed data sources. Results: Here, we describe the development of ‘dsMTL’, a computational framework for privacy-preserving, distributed multi-task machine learning that includes three supervised and one unsupervised algorithms. First, we derive the theoretical properties of these methods and the relevant machine learning workflows to ensure the validity of the software implementation. Second, we implement dsMTL as a library for the R programming language, building on the DataSHIELD platform that supports the federated analysis of sensitive individual-level data. Third, we demonstrate the applicability of dsMTL for comorbidity modeling in distributed data. We show that comorbidity modeling using dsMTL outperformed conventional, federated machine learning, as well as the aggregation of multiple models built on the distributed datasets individually. The application of dsMTL was computationally efficient and highly scalable when applied to moderate-size (n < 500), real expression data given the actual network latency. Availability and implementation: dsMTL is freely available at https://github.com/transbioZI/dsMTLBase (server-side package) and https://github.com/transbioZI/dsMTLClient (client-side package).
UR - http://www.scopus.com/inward/record.url?scp=85141005634&partnerID=8YFLogxK
U2 - 10.1093/bioinformatics/btac616
DO - 10.1093/bioinformatics/btac616
M3 - Article
C2 - 36073911
AN - SCOPUS:85141005634
SN - 1367-4803
VL - 38
SP - 4919
EP - 4926
JO - Bioinformatics
JF - Bioinformatics
IS - 21
ER -