Valid inference for machine learning-assisted genome-wide association studies
Open accessPublished version of a preprint
Abstract
Machine learning (ML) has become increasingly popular in almost all scientific disciplines, including human genetics. Owing to challenges related to sample collection and precise phenotyping, ML-assisted genome-wide association study (GWAS), which uses sophisticated ML techniques to impute phenotypes and then performs GWAS on the imputed outcomes, have become increasingly common in complex trait genetics research. However, the validity of ML-assisted GWAS associations has not been carefully evaluated. Here, we report pervasive risks for false-positive associations in ML-assisted GWAS and introduce Post-Prediction GWAS (POP-GWAS), a statistical framework that redesigns GWAS on ML-imputed outcomes. POP-GWAS ensures valid and powerful statistical inference irrespective of imputation quality and choice of algorithm, requiring only GWAS summary statistics as input. We employed POP-GWAS to perform a GWAS of bone mineral density derived from dual-energy X-ray absorptiometry imaging at 14 skeletal sites, identifying 89 new loci and revealing skeletal site-specific genetic architecture. Our framework offers a robust analytic solution for future ML-assisted GWAS.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · see citing works
Highly accurate protein structure prediction with AlphaFold
Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Žídek, A., Potapenko, A., et al. (2021) · Nature · doi:10.1038/s41586-021-03819-2 · Cited by 43,764 · see citing works
Opportunities and challenges in long-read sequencing data analysis
Amarasinghe, S. L., Su, S., Dong, X., Zappia, L., Ritchie, M. E., & Gouil, Q. (2020) · Genome Biology · doi:10.1186/s13059-020-1935-5 · Cited by 2,530 · see citing works
Genome-wide association studies
Uffelmann, E., Huang, Q. Q., Munung, N. S., de Vries, J., Okada, Y., Martin, A. R., Martin, H. C., Lappalainen, T., & Posthuma, D. (2021) · Nature Reviews Methods Primers · doi:10.1038/s43586-021-00056-9 · Cited by 1,849 · see citing works
Plasma proteomic associations with genetics and health in the UK Biobank
Sun, B. B., Chiou, J., Traylor, M., Benner, C., Hsu, Y.-H., Richardson, T. G., Surendran, P., Mahajan, A., Robins, C., Vasquez-Grinnell, S. G., et al. (2023) · Nature · doi:10.1038/s41586-023-06592-6 · Cited by 1,740 · see citing works
Multi-trait analysis of genome-wide association summary statistics using MTAG
23andMe Research Team, Turley, P., Social Science Genetic Association Consortium, Walters, R. K., Maghzian, O., Okbay, A., Lee, J. J., Fontana, M. A., Nguyen-Viet, T. A., Wedow, R., et al. (2018) · Nature Genetics · doi:10.1038/s41588-017-0009-4 · Cited by 1,112 · see citing works
Eliciting priors and relaxing the single causal variant assumption in colocalisation analyses
Wallace, C. (2020) · PLOS Genetics · doi:10.1371/journal.pgen.1008720 · Cited by 441 · see citing works
Wheeler, E., Leong, A., Liu, C.-T., Hivert, M.-F., Strawbridge, R. J., Podmore, C., Li, M., Yao, J., Sim, X., Hong, J., et al. (2017) · PLOS Medicine · doi:10.1371/journal.pmed.1002383 · Cited by 408 · see citing works
Lu, Q., Powles, R. L., Abdallah, S., Ou, D., Wang, Q., Hu, Y., Lu, Y., Liu, W., Li, B., Mukherjee, S., et al. (2017) · PLOS Genetics · doi:10.1371/journal.pgen.1006933 · Cited by 97 · see citing works
Self-citationImputing Phenotypes for Genome-wide Association Studies
Hormozdiari, F., Kang, E. Y., Bilow, M., Ben-David, E., Vulpe, C., McLachlan, S., Lusis, A. J., Han, B., & Eskin, E. (2016) · The American Journal of Human Genetics · doi:10.1016/j.ajhg.2016.04.013 · Cited by 53 · see citing works
Flagged first, then by citations
- Authors
- Miao, J., Wu, Y., Sun, Z., Miao, X., Lu, T., Zhao, J., & Lu, Q.
- Affiliations
- University of Wisconsin–Madison (US)
- University of Toronto (CA)
- Corresponding author
- Lu, Qiongshi
- Venue
- Nature Genetics
- Published
- 2024-09-30
- Volume
- 56
- Issue
- 11
- Pages
- 2361-2369
- ISSN
- 1061-4036 (ISSN-L)
- 1546-1718
- PMC
- PMC11972620
- OpenAlex
- W4402992718
- Cited by
- 28
- References
- 10
- Citations by year
- 2026: 5 · 2025: 20 · 2024: 3
- FWCI
- 7.43
- Type
- journal article
- Language
- en
- Open access
- green
- doi.org
- Licence
- Locations
- Nature Genetics (publishedVersion)
- medRxiv (acceptedVersion, cc-by-nc-nd)
- PubMed (publishedVersion)
- Europe PMC (PubMed Central) (acceptedVersion)
- PubMed Central (acceptedVersion)
- Funders
- U.S. Department of Health & Human Services | National Institutes of Health (U01 HG012039)
- U.S. Department of Health & Human Services | NIH | National Institute on Aging (P30 AG017266)
- Topics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Associations and Epidemiology
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Mapping and Diversity in Plants and Animals
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic and phenotypic traits in livestock
- Keywords
- false positive association, bone mineral density, complex trait genetics, statistical inference validity
- MeSH terms
- Machine Learning, Algorithms, Humans, Phenotype, Bone Density, Polymorphism, Single Nucleotide, Genome-Wide Association Study
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
