Leveraging functional annotations in genetic risk prediction for human complex diseases

Hu, Yiming; Lu, Qiongshi; Powles, Ryan L.; Yao, Xinwei; Yang, Can; Fang, Fang; Xu, Xinran; Zhao, Hongyu

doi:10.1371/journal.pcbi.1005589

Cited by 162 publications

(139 citation statements)

References 35 publications

Supporting

Mentioning

133

Contrasting

Order By: Relevance

“…Extended Data Fig. 1 confirms that when we run our versions assuming the GCTA Model, they perform at least as as well, both in terms of prediction accuracy and computational efficiency, as existing versions (namely the original versions of Bolt-LMM and BayesR, 13,14 both of which use individual-level data, and the summary statistic tools lassosum, sBLUP, LDPred, AnnoPred and SBayesR 15,17,[20][21][22][23] ). Additionally, we develop a new summary statistics tool, MegaPRS, which constructs lasso, ridge regression, Bolt-LMM and BayesR models, then selects the best-performing tool via cross-validation (it does this at the same time as it selects prior parameters for each tool).…”

Section: Resultssupporting

confidence: 55%

“…We are aware of two summary summary statistic prediction tools where the user can specify the heritability model, AnnoPred and LDPred-funct. 22,23 AnnoPred is similar to Bolt-LMM. It assumes that SNP effect sizes have the prior distribution p 0 N(0,σ) 2 ) + (1-p 0 ) δ 0 (this matches the Bolt-LMM prior distribution when σ) 2 Small =0), then incorporates the chosen heritability model by allowing either σ) 2 or p 0 to vary across SNPs.…”

Section: Discussionmentioning

confidence: 95%

See 1 more Smart Citation

Improved genetic prediction of complex traits from individual-level data or summary statistics

Zhang

Privé

Vilhjálmsson

et al. 2020

Preprint

View full text Add to dashboard Cite

At present, most tools for constructing genetic prediction models begin with the assumption that all genetic variants contribute equally towards the phenotype. However, this represents a sub-optimal model for how heritability is distributed across the genome. Here we construct prediction models for 14 phenotypes from the UK Biobank (200,000 individuals per phenotype) using four of the most popular prediction tools: lasso, ridge regression, Bolt-LMM and BayesR. When we improve the assumed heritability model, prediction accuracy always improves (i.e., for all four tools and for all 14 phenotypes). When we construct prediction models using individual-level data, the best-performing tool is Bolt-LMM; if we replace its default heritability model with the most realistic model currently available, the average proportion of phenotypic variance explained increases by 19% (s.d. 2), equivalent to increasing the sample size by about a quarter. When we construct prediction models using summary statistics, the best tool depends on the phenotype. Therefore, we develop MegaPRS, a summary statistic prediction tool for constructing lasso, ridge regression, Bolt-LMM and BayesR prediction models, that allows the user to specify the heritability model.

show abstract

Section: Resultssupporting

confidence: 55%

Section: Discussionmentioning

confidence: 95%

Improved genetic prediction of complex traits from individual-level data or summary statistics

Zhang

Privé

Vilhjálmsson

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…To further improve prediction accuracy, several methods have been proposed to utilize other information, such as LDpred (and LDpred-inf) that models the LD information extracted from a reference panel [6]; AnnoPred that leverages diverse types of genomic and epigenomic functional annotations [7]; and PleioPred and SMTpred that utilize pleiotropy relationship with other traits/diseases [8,9]. All of these methods need to borrow information from external panels or datasets.…”

Section: Introductionmentioning

confidence: 99%

Leveraging effect size distributions to improve polygenic risk scores derived from summary statistics of genome-wide association studies

et al. 2020

Self Cite

View full text Add to dashboard Cite

Genetic risk prediction is an important problem in human genetics, and accurate prediction can facilitate disease prevention and treatment. Calculating polygenic risk score (PRS) has become widely used due to its simplicity and effectiveness, where only summary statistics from genome-wide association studies are needed in the standard method. Recently, several methods have been proposed to improve standard PRS by utilizing external information, such as linkage disequilibrium and functional annotations. In this paper, we introduce EB-PRS, a novel method that leverages information for effect sizes across all the markers to improve prediction accuracy. Compared to most existing genetic risk prediction methods, our method does not need to tune parameters nor external information. Real data applications on six diseases, including asthma, breast cancer, celiac disease, Crohn's disease, Parkinson's disease and type 2 diabetes show that EB-PRS achieved 307.1%, 42.8%, 25.5%, 3.1%, 74.3% and 49.6% relative improvements in terms of predictive r 2 over standard PRS method with optimally tuned parameters. Besides, compared to LDpred that makes use of LD information, EB-PRS also achieved 37.9%, 33.6%, 8.6%, 36.2%, 40.6% and 10.8% relative improvements. We note that our method is not the first method leveraging effect size distributions. Here we first justify our method by presenting theoretical optimal property over existing methods in this class of methods, and substantiate our theoretical result with extensive simulation results. The R-package EBPRS that implements our method is available on CRAN. Author summaryGenetic risk prediction is of considerable importance in human genetics. Calculating polygenic risk score (PRS) is the most commonly used approach due to its simplicity as well as the computational efficiency. Many PRS calculation methods have been proposed for accurate prediction by borrowing information from external panels or datasets. Here, instead of external information, we propose a novel PRS method leveraging internal PLOS Computational Biology | https://doi.

show abstract

“…As shown in [9, 13], LDpred and AnnoPred outperform other state-of-the-art PRS methods, we therefore use these two approaches as the representative single-trait prediction methods.…”

Section: Resultsmentioning

confidence: 99%

Joint modeling of genetically correlated diseases and functional annotations increases accuracy of polygenic risk prediction

Liu

et al. 2017

PLoS Genet

Self Cite

View full text Add to dashboard Cite

Accurate prediction of disease risk based on genetic factors is an important goal in human genetics research and precision medicine. Advanced prediction models will lead to more effective disease prevention and treatment strategies. Despite the identification of thousands of disease-associated genetic variants through genome-wide association studies (GWAS) in the past decade, accuracy of genetic risk prediction remains moderate for most diseases, which is largely due to the challenges in both identifying all the functionally relevant variants and accurately estimating their effect sizes. In this work, we introduce PleioPred, a principled framework that leverages pleiotropy and functional annotations in genetic risk prediction for complex diseases. PleioPred uses GWAS summary statistics as its input, and jointly models multiple genetically correlated diseases and a variety of external information including linkage disequilibrium and diverse functional annotations to increase the accuracy of risk prediction. Through comprehensive simulations and real data analyses on Crohn’s disease, celiac disease and type-II diabetes, we demonstrate that our approach can substantially increase the accuracy of polygenic risk prediction and risk population stratification, i.e. PleioPred can significantly better separate type-II diabetes patients with early and late onset ages, illustrating its potential clinical application. Furthermore, we show that the increment in prediction accuracy is significantly correlated with the genetic correlation between the predicted and jointly modeled diseases.

show abstract

Leveraging functional annotations in genetic risk prediction for human complex diseases

Cited by 162 publications

References 35 publications

Improved genetic prediction of complex traits from individual-level data or summary statistics

Improved genetic prediction of complex traits from individual-level data or summary statistics

Leveraging effect size distributions to improve polygenic risk scores derived from summary statistics of genome-wide association studies

Joint modeling of genetically correlated diseases and functional annotations increases accuracy of polygenic risk prediction

Contact Info

Product

Resources

About