| Title: | A Selection Operator for Summary Association Statistics Reveals Allelic Heterogeneity of Complex Traits |
| Journal: | American Journal of Human Genetics |
| Published: | 5 Dec 2017 |
| Pubmed: | https://pubmed.ncbi.nlm.nih.gov/29198721/ |
| DOI: | https://doi.org/10.1016/j.ajhg.2017.09.027 |
| URL: | http://www.cell.com/article/S0002929717304196/pdf |
| Citations: | 8 as of 8 Aug 2024 |
Publication 3944
WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
Abstract
In recent years, as a secondary analysis in genome-wide association studies (GWASs), conditional and joint multiple-SNP analysis (GCTA-COJO) has been successful in allowing the discovery of additional association signals within detected loci. This suggests that many loci mapped in GWASs harbor more than a single causal variant. In order to interpret the underlying mechanism regulating a complex trait of interest in each discovered locus, researchers must assess the magnitude of allelic heterogeneity within the locus. We developed a penalized selection operator for jointly analyzing multiple variants (SOJO) within each mapped locus on the basis of LASSO (least absolute shrinkage and selection operator) regression derived from summary association statistics. We found that, compared to stepwise conditional multiple-SNP analysis, SOJO provided better sensitivity and specificity in predicting the number of alleles associated with complex traits in each locus. SOJO suggested causal variants potentially missed by GCTA-COJO. Compared to using top variants from genome-wide significant loci in GWAS, using SOJO increased the proportion of variance prediction for height by 65% without additional discovery samples or additional loci in the genome. Our empirical results indicate that human height is not only a highly polygenic trait, but also has high allelic heterogeneity within its established hundreds of loci.6 Authors
- Zheng Ning
- Youngjo Lee
- Peter K. Joshi
- James F. Wilson
- Yudi Pawitan
- Xia Shen
Enabling scientific discoveries that improve human health