Federated Generalized Linear Mixed Models for Collaborative Genome-Wide Association Studies

Li, Wentao; Chen, Han; Jiang, Xiaoqian; Harmanci, Arif

doi:10.2139/ssrn.4211802

Cited by 3 publications

(2 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This mechanism can provide further protection against haplotype decoding by increasing the local haplotype distribution entropy beyond the mechanism that we used in this study. We foresee that new mechanisms can be designed for other tasks, e.g., kinship estimation [87], collaborative GWAS [88], as well. These mechanisms can be combined with encryption-based techniques to decrease further privacy risks while increasing efficiency.…”

Section: Discussionmentioning

confidence: 99%

ProxyTyper: Generation of Proxy Panels for Privacy-aware Outsourcing of Genotype Imputation

Zhi,

Jiang,

Harmanci

2023

Preprint

View full text Add to dashboard Cite

One of the major challenges in genomic data sharing is protecting the privacy of participants. Numerous studies demonstrated that genetic data and related summary statistics can be used for identifying individuals. These studies led to a strong chilling effect on researchers that hindered collaborative data sharing. Consequently, population-level genetic databases are often siloed in central repositories with complex and burdensome data usage agreements. While cryptographic methods that are provably secure have been developed, these methods require high-level expertise in security and depend on large computational resources.To fill the methodological gap in this domain, we present ProxyTyper, a set of data protection mechanisms to generate “proxy-panels” from sensitive haplotype-level genetic datasets. ProxyTyper uses haplotype sampling, allele hashing, and anonymization to protect the genetic variant coordinates, genetic maps, and chromosome-wide haplotypes. These mechanisms can provide strong deterrence against honest-but-curious entities and well-known re-identification and linking attacks. The proxy panels can be used as input to existing tools without any modifications to the underlying algorithms. We focus on building proxy-panels for genotype imputation to protect typed and untyped variants. We demonstrate that proxy-based imputation provides protection against well-known attacks with a minor decrease of imputation accuracy for variants in wide range of allele frequencies.

show abstract

Section: Discussionmentioning

confidence: 99%

ProxyTyper: Generation of Proxy Panels for Privacy-aware Outsourcing of Genotype Imputation

Zhi,

Jiang,

Harmanci

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Phenotype information was simulated evaluating a logistic link function on the weighted linear combination of covariates and genetic effects on each individual. Population covariates for both homogeneous and heterogeneous datasets were estimated by projection of the genotype data onto 1000 Genomes reference panel [46]. For heterogeneous sample, the genotype data was generated from 3 populations (GBR, YRI, MXL).…”

Section: B Experimental Setupmentioning

confidence: 99%

FedGMMAT: Federated Generalized Linear Mixed Model Association Tests

Li,

Chen,

Jiang

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

Increasing genetic and phenotypic data size is critical for understanding the genetic determinants of diseases. Evidently, establishing practical means for collaboration and data sharing among institutions is a fundamental methodological barrier for performing high-powered studies. As the sample sizes become more heterogeneous, complex statistical approaches, such as generalized linear mixed effects models, must be used to correct for confounders that may bias results. On another front, due to the privacy concerns around Protected Health Information (PHI), genetic information is restrictively protected by sharing according to regulations such as Health Insurance Portability and Accountability Act (HIPAA). This limits data sharing among institutions and hampers efforts around executing high-powered collaborative studies. Federated approaches are promising to alleviate the issues around privacy and performance, since sensitive data never leaves the local sites.Motivated by these, we developed FedGMMAT, a federated genetic association testing tool that utilizes a federated statistical testing approach for efficient association tests that can correct for arbitrary fixed and random effects among different collaborating sites. Genetic data is never shared among collaborating sites, and the intermediate statistics are protected by homomorphic encryption. Using simulated and real datasets, we demonstrate FedGMMAT can achieve the virtually same results as pooled analysis under a privacy-preserving framework with practical resource requirements.

show abstract

FedGMMAT: Federated generalized linear mixed model association tests

Li,

Chen,

Jiang

et al. 2024

PLoS Comput Biol

View full text Add to dashboard Cite

Increasing genetic and phenotypic data size is critical for understanding the genetic determinants of diseases. Evidently, establishing practical means for collaboration and data sharing among institutions is a fundamental methodological barrier for performing high-powered studies. As the sample sizes become more heterogeneous, complex statistical approaches, such as generalized linear mixed effects models, must be used to correct for the confounders that may bias results. On another front, due to the privacy concerns around Protected Health Information (PHI), genetic information is restrictively protected by sharing according to regulations such as Health Insurance Portability and Accountability Act (HIPAA). This limits data sharing among institutions and hampers efforts around executing high-powered collaborative studies. Federated approaches are promising to alleviate the issues around privacy and performance, since sensitive data never leaves the local sites. Motivated by these, we developed FedGMMAT, a federated genetic association testing tool that utilizes a federated statistical testing approach for efficient association tests that can correct for confounding fixed and additive polygenic random effects among different collaborating sites. Genetic data is never shared among collaborating sites, and the intermediate statistics are protected by encryption. Using simulated and real datasets, we demonstrate FedGMMAT can achieve the virtually same results as pooled analysis under a privacy-preserving framework with practical resource requirements.

show abstract

Federated Generalized Linear Mixed Models for Collaborative Genome-Wide Association Studies

Cited by 3 publications

References 41 publications

ProxyTyper: Generation of Proxy Panels for Privacy-aware Outsourcing of Genotype Imputation

ProxyTyper: Generation of Proxy Panels for Privacy-aware Outsourcing of Genotype Imputation

FedGMMAT: Federated Generalized Linear Mixed Model Association Tests

FedGMMAT: Federated generalized linear mixed model association tests

Contact Info

Product

Resources

About