Efficacy of federated learning on genomic data: a study on the UK Biobank and the 1000 Genomes Project

Abstract

Combining training data from multiple sources increases sample size and reduces confounding, leading to more accurate and less biased machine learning models. In healthcare, however, direct pooling of data is often not allowed by data custodians who are accountable for minimizing the exposure of sensitive information. Federated learning offers a promising solution to this problem by training a model in a decentralized manner thus reducing the risks of data leakage. Although there is increasing utilization of federated learning on clinical data, its efficacy on genomic data has not been extensively studied. This study aims to contribute to the adoption of federated learning for genomic data by investigating its applicability in two scenarios: phenotype prediction on the UK Biobank data and ancestry prediction on the 1000 Genomes Project data. By splitting data into independent datasets and implementing different federated learning strategies, we show that federated models achieve performance close to centralized models, even in the presence of significant inter-node heterogeneity. This paper describes the experiments and provides recommendations on strategies that should be used to reduce computational complexity or communication costs.

An Integrative Pancreatic Cancer Risk Prediction Model in the UK Biobank

“…Linkage disequilibrium (LD) was also assessed using the threshold for LD pruning (r 2 > 0.8) (excluding 4 SNPs, n = 68 SNPs). This study did not include any ambiguous, palindromic SNPs or SNPs with an imputation score < 0.8 [36]. Finally, only the SNPs that passed the SNP QC from the Caucasian GWAS (n = 40 SNPs) were included in this study.…”