Computationally efficient whole-genome regression for quantitative and binary traits

Mbatchou, Joelle; Barnard, Leland; Backman, Joshua; Marcketta, Anthony; Kosmicki, Jack A.; Ziyatdinov, Andrey; Benner, Christian; Ó'Dúshláine, Colm; Barber, Mathew; Boutkov, Boris; Habegger, Lukas; Ferreira, Manuel A. R.; Baras, Aris; Reid, Jeffrey G.; Abecasis, Gonçalo R.; Maxwell, Evan K.; Marchini, Jonathan

doi:10.1038/s41588-021-00870-7

Cited by 800 publications

(703 citation statements)

References 39 publications

Supporting

Mentioning

697

Contrasting

Unclassified

Order By: Relevance

“…Phenotypes were adjusted for covariates and inverse normalized, as described above for AA and HL cohorts (additional details in Supplemental Content, Included Cohorts). REGENIE [40] was used to test the association between predicted gene expression and adjusted phenotypes.…”

Section: Replicationmentioning

confidence: 99%

Transcriptome-Wide Association Study of Blood Cell Traits in African Ancestry and Hispanic/Latino Populations

Wen

Xie

Rowland

et al. 2021

Genes

View full text Add to dashboard Cite

Background: Thousands of genetic variants have been associated with hematological traits, though target genes remain unknown at most loci. Moreover, limited analyses have been conducted in African ancestry and Hispanic/Latino populations; hematological trait associated variants more common in these populations have likely been missed. Methods: To derive gene expression prediction models, we used ancestry-stratified datasets from the Multi-Ethnic Study of Atherosclerosis (MESA, including n = 229 African American and n = 381 Hispanic/Latino participants, monocytes) and the Depression Genes and Networks study (DGN, n = 922 European ancestry participants, whole blood). We then performed a transcriptome-wide association study (TWAS) for platelet count, hemoglobin, hematocrit, and white blood cell count in African (n = 27,955) and Hispanic/Latino (n = 28,324) ancestry participants. Results: Our results revealed 24 suggestive signals (p < 1 × 10−4) that were conditionally distinct from known GWAS identified variants and successfully replicated these signals in European ancestry subjects from UK Biobank. We found modestly improved correlation of predicted and measured gene expression in an independent African American cohort (the Genetic Epidemiology Network of Arteriopathy (GENOA) study (n = 802), lymphoblastoid cell lines) using the larger DGN reference panel; however, some genes were well predicted using MESA but not DGN. Conclusions: These analyses demonstrate the importance of performing TWAS and other genetic analyses across diverse populations and of balancing sample size and ancestry background matching when selecting a TWAS reference panel.

show abstract

Section: Replicationmentioning

confidence: 99%

Transcriptome-Wide Association Study of Blood Cell Traits in African Ancestry and Hispanic/Latino Populations

Wen

Xie

Rowland

et al. 2021

Genes

View full text Add to dashboard Cite

show abstract

“…We defined genome-wide significant loci in our analysis by linkage disequilibrium (r 2 > 0.1) with lead variants. We defined previously associated loci by their index variants reported in previous hearing loss GWAS [12][13][14][15][16]56,[69][70][71][72][73] , and excluded 1 Mb regions surrounding them in the identification of novel loci in our analysis. LD score (LDSC) regression 75 was used to assess inflation (LDSC intercept) accounting for polygenic signal.…”

Section: Fine Mapping and Follow-on Genetic Analysesmentioning

confidence: 99%

Population-scale analysis of common and rare genetic variation associated with hearing loss in adults

Praveen

Dobbyn

Gurski

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Understanding the genetic underpinnings of disabling hearing loss, which affects ~466 million people worldwide, can provide avenues for new therapeutic target development. We performed a genome-wide association meta-analysis of hearing loss with 125,749 cases and 469,497 controls across five cohorts, including UK Biobank, Geisinger DiscovEHR, the Malmö Diet and Cancer Study, the Mount Sinai BioMe Personalized Medicine Cohort, and FinnGen. We identified 53 loci affecting hearing loss risk, 18 of which are novel, including coding variants in COL9A3 and TMPRSS3. Through exome-sequencing of 108,415 cases and 329,581 controls from the same cohorts, we identified an association with rare predicted loss-of-function variants in a gene that has been uncharacterized in hearing loss, KLHDC7B (odds ratio [OR] = 2.14, P = 5.2 × 10-30), and with coding variants in two genes previously implicated in animal models of hearing loss (SYNJ2, OR = 1.31, P = 1.3 × 10-14; FSCN2, OR = 1.24, P = 4.1 × 10-15). We also observed single-variant and gene-burden associations with 11 genes known to cause Mendelian forms of hearing loss, including an increased risk in heterozygous carriers of mutations in the autosomal recessive hearing loss genes GJB2 (Gly12fs; OR = 1.21, P = 4.2 × 10-11) and SLC26A5 (gene burden; OR = 1.96, P = 2.8 × 10-17). Our results show that Mendelian hearing loss genes contribute to the burden of hearing loss in the adult population, and suggest a shared etiology between common and rare forms of hearing loss. This work illustrates the potential of large-scale exome sequencing to elucidate the genetic architecture of common traits in which risk is modulated by both common and rare variation.

show abstract

“…Recently, Rounak et al proposed a fast genotype odds ratios estimation in which Firth’s penalty was adjusted ( Dey and Lee, 2019 ). REGENIE also used an approximate Firth regression in which covariate effects were incorporated through an offset term ( Mbatchou et al, 2021 ). These strategies reduce the number of predictors and thus are scalable in GWAS.…”

Section: Statistical and Computational Challenges In Biobank Data Analysis And Approaches To Addressing Themmentioning

confidence: 99%

“…Instead of the mixed effect model framework, a fixed effect model with a penalty can be used to account for genetic relatedness. A recent developed REGENIE ( Mbatchou et al, 2021 ) used two-step ridge regressions to calculate predictors from genetic data and then used linear and logistic regression to associate quantitative and binary traits with genetic variants. Compared to mixed effect model approaches, fixed effect model approaches can be faster and needs a smaller amount of memory.…”

Section: Statistical and Computational Challenges In Biobank Data Analysis And Approaches To Addressing Themmentioning

confidence: 99%

Scalable and Robust Regression Methods for Phenome-Wide Association Analysis on Large-Scale Biobank Data

Lee

2021

Front. Genet.

View full text Add to dashboard Cite

With the advances in genotyping technologies and electronic health records (EHRs), large biobanks have been great resources to identify novel genetic associations and gene-environment interactions on a genome-wide and even a phenome-wide scale. To date, several phenome-wide association studies (PheWAS) have been performed on biobank data, which provides comprehensive insights into many aspects of human genetics and biology. Although inspiring, PheWAS on large-scale biobank data encounters new challenges including computational burden, unbalanced phenotypic distribution, and genetic relationship. In this paper, we first discuss these new challenges and their potential impact on data analysis. Then, we summarize approaches that are scalable and robust in GWAS and PheWAS. This review can serve as a practical guide for geneticists, epidemiologists, and other medical researchers to identify genetic variations associated with health-related phenotypes in large-scale biobank data analysis. Meanwhile, it can also help statisticians to gain a comprehensive and up-to-date understanding of the current technical tool development.

show abstract

Computationally efficient whole-genome regression for quantitative and binary traits

Cited by 800 publications

References 39 publications

Transcriptome-Wide Association Study of Blood Cell Traits in African Ancestry and Hispanic/Latino Populations

Transcriptome-Wide Association Study of Blood Cell Traits in African Ancestry and Hispanic/Latino Populations

Population-scale analysis of common and rare genetic variation associated with hearing loss in adults

Scalable and Robust Regression Methods for Phenome-Wide Association Analysis on Large-Scale Biobank Data

Contact Info

Product

Resources

About