Version: 1.0.0
Authors: Dovini Jayasinghe and S. Hong Lee
The iPRSue is a novel method for estimating polygenic risk scores (PRSs) and their uncertainty (variance) using Genome Wide Association Studies (GWAS) summary statistics which can be constructed by GWAS_QT() or GWAS_BT() functions (see the section iPRSue_estimates_QT() and iPRSue_estimates_BT(), providing unbiased and precise estimates. In addition, we have incorporated traditionally used Best linear unbiased estimator (BLUE) method for PRS and variance estimation, available for both trait types via BLUE_estimates_QT() and BLUE_estimates_BT(). While the BLUE method offers slightly but significantly higher prediction accuracy compared to iPRSue, it is limited to cases where the number of SNP genotypes is smaller than the discovery sample size. In contrast, iPRSue does not have this limitation and can be applied to any data dimension.
Install devtools:
install.packages("devtools")
Install iPRSue:
devtools::install_github("DoviniJ/iPRSue")
library(iPRSue)
- Qpd.txt - This is a .txt file which contains the following columns in order. The discovery dataset has 800 individuals. Note that the file has no column headings.
- family ID (FID)
- individual ID (IID)
- quantitative phenotype of the discovery sample
ID_1 ID_1 31.6534
ID_2 ID_2 25.5035
ID_3 ID_3 26.7391
ID_4 ID_4 25.5271
ID_5 ID_5 26.7165
- Qpt.txt - This is a .txt file which contains the following columns in order. The target dataset has 200 individuals who are independent from the discovery dataset. Note that the file has no column headings.
- FID
- IID
- quantitative phenotype of the target sample - optional
ID_801 ID_801 26.5723
ID_802 ID_802 20.2632
ID_803 ID_803 27.7365
ID_804 ID_804 18.75
ID_805 ID_805 23.3025
- Bpd.txt - This is a .txt file which contains the following columns in order. The discovery dataset has 800 individuals. Note that the file has no column headings.
- FID
- IID
- binary phenotype (0=controls, 1=cases) of the discovery sample
ID_1 ID_1 0
ID_2 ID_2 1
ID_3 ID_3 0
ID_4 ID_4 0
ID_5 ID_5 0
- Bpt.txt - This is a .txt file which contains the following columns in order. The target dataset has 200 individuals who are independent from the discovery dataset. Note that the file has no column headings.
- FID
- IID
- binary phenotype (0=controls, 1=cases) of the target sample - optional
ID_801 ID_801 0
ID_802 ID_802 1
ID_803 ID_803 0
ID_804 ID_804 0
ID_805 ID_805 1
-
Gd.txt - This is a .txt file which contains scaled (column standardized) genotype matrix with the dimension discovery_sample_size x number_of_snps (e.g. 800 x 100), of the discovery individuals. Note that the file has neither row nor column headings.
-
Gt.txt - This is a .txt file which contains scaled (column standardized) genotype matrix with the dimension target_sample_size x number_of_snps (e.g. 200 x 100), of the target individuals. Note that the file has neither row nor column headings.
Download all the above example data files from inst folder to your working directory. Follow the commands below to obtain the expected outputs. Once practiced, you may use your own data files to generate individual PRS confidence intervals by using either the novel method iPRSue or the BLUE method.
Commands
x <- GWAS_QT(discovery_pheno = "Qpd.txt", discovery_geno_mat = "Gd.txt")
y <- iPRSue_estimates_QT(gwas = x, target_pheno = "Qpt.txt", target_geno_mat = "Gt.txt", no_of_PRSs = 500, significance_level = 0.05, seed = set.seed(1))
The arguments of GWAS_QT() function, namely, discovery_pheno and discovery_geno_mat should be the text file names of phenotype and scaled genotype matrix for the individuals in the discovery dataset. The function iPRSue_estimates() use the estimated values from GWAS_QT() function as an input to the argument gwas. Moreover, target_pheno and target_geno_mat should be the text file names of phenotype and scaled genotype matrix for the individuals in the target dataset. The no_of_PRSs specifies the size of the PRS distribution per target individual (default is 500), significance_level sets the level of significance for constructing PRS confidence intervals (default is 0.05), and seed is used to control random number generation for reproducibility (default is NULL).
Outputs
beta se
1 -0.005625531 0.03539906
2 0.011408607 0.03539731
3 -0.024074395 0.03538936
4 -0.021290422 0.03539159
5 0.032671483 0.03538072
6 0.019838674 0.03539265
Returns the dataframe x with following columns:
- x$beta : Additive effects of scaled SNP genotypes
- x$se : Standard errors of the additive effects
IID PRS Variance Lower_Limit Upper_Limit
1 ID_801 -0.10518765 0.1231988 -0.7764618 0.5997384
2 ID_802 -0.34199622 0.1075602 -0.9481694 0.2861330
3 ID_803 -0.35311776 0.1041257 -0.9409285 0.2779039
4 ID_804 0.07428962 0.1257936 -0.6987473 0.6993487
5 ID_805 -0.03311903 0.1191919 -0.7008964 0.6027532
6 ID_806 -0.04271522 0.1294443 -0.7324805 0.6695980
Returns the dataframe y with following columns:
- y$IID : Target individual IDs
- y$PRS : PRS estimates of each target individual, computed using iPRSue method
- y$Variance : Variance of individual PRS, computed using iPRSue method
- y$Lower_Limit : Lower limit of the individual PRS confidence interval
- y$Upper_Limit : Lower limit of the individual PRS confidence interval
Command
z <- BLUE_estimates_QT(discovery_pheno = "Qpd.txt", discovery_geno_mat = "Gd.txt", target_pheno = "Qpt.txt", target_geno_mat = "Gt.txt", significance_level = 0.05)
The function BLUE_estimates_QT() utilizes individual level data and provides PRS and uncertainty estimates using the BLUE multiple linear regression approach.
Output
IID PRS Variance Lower_Limit Upper_Limit
1 ID_801 -0.1444712318 0.1360387 -0.8673731 0.5784306
2 ID_802 -0.3937930417 0.1255455 -1.0882554 0.3006693
3 ID_803 -0.5239819895 0.1410622 -1.2601100 0.2121460
4 ID_804 0.1379861095 0.1386364 -0.5917851 0.8677573
5 ID_805 -0.0009677077 0.1527249 -0.7669223 0.7649869
6 ID_806 -0.2428045632 0.1490595 -0.9995118 0.5139027
Returns the dataframe z with following columns:
- y$IID : Target individual IDs
- y$PRS : PRS estimates of each target individual, computed using BLUE method
- y$Variance : Variance of individual PRS, computed using BLUE method
- y$Lower_Limit : Lower limit of the individual PRS confidence interval
- y$Upper_Limit : Lower limit of the individual PRS confidence interval
Commands
x <- GWAS_BT(discovery_pheno = "Bpd.txt", discovery_geno_mat = "Gd.txt")
y <- iPRSue_estimates_BT(gwas = x, target_pheno = "Bpt.txt", target_geno_mat = "Gt.txt", no_of_PRSs = 500, significance_level = 0.05, seed = set.seed(1))
The arguments of GWAS_BT() function, namely, discovery_pheno and discovery_geno_mat should be the text file names of phenotype and scaled genotype matrix for the individuals in the discovery dataset. The function iPRSue_estimates() use the estimated values from GWAS_BT() function as an input to the argument gwas. Moreover, target_pheno and target_geno_mat should be the text file names of phenotype and scaled genotype matrix for the individuals in the target dataset. The no_of_PRSs specifies the size of the PRS distribution per target individual (default is 500), significance_level sets the level of significance for constructing PRS confidence intervals (default is 0.05), and seed is used to control random number generation for reproducibility (default is NULL).
Outputs
beta se
1 0.02351554 0.1734839
2 0.19951530 0.1609413
3 0.01422655 0.1742960
4 -0.03176976 0.1777684
5 -0.18495543 0.1921530
6 -0.10750075 0.1847582
Returns the dataframe x with following columns:
- x$beta : Additive effects of scaled SNP genotypes
- x$se : Standard errors of the additive effects
IID PRS Variance Lower_Limit Upper_Limit
1 ID_801 2.876907 3.185277 -0.4678476 6.5059090
2 ID_802 1.802006 2.769852 -1.3031663 4.9981202
3 ID_803 2.431920 2.697261 -0.4885771 5.6501721
4 ID_804 -2.524441 3.273875 -6.4328362 0.7284884
5 ID_805 2.363432 3.063884 -1.1095964 5.6116642
6 ID_806 1.497236 3.322266 -2.0080071 5.1538390
Returns the dataframe y with following columns:
- y$IID : Target individual IDs
- y$PRS : PRS estimates of each target individual, computed using iPRSue method
- y$Variance : Variance of individual PRS, computed using iPRSue method
- y$Lower_Limit : Lower limit of the individual PRS confidence interval
- y$Upper_Limit : Lower limit of the individual PRS confidence interval
Command
z <- BLUE_estimates_BT(discovery_pheno = "Bpd.txt", discovery_geno_mat = "Gd.txt", target_pheno = "Bpt.txt", target_geno_mat = "Gt.txt", significance_level = 0.05, max_iterations = 100)
The function BLUE_estimates_BT() utilizes individual level data and provides PRS and uncertainty estimates using the BLUE multiple logistic regression approach with firth's bias correction.
Output
IID PRS Variance Lower_Limit Upper_Limit
1 ID_801 2.701390 2.194493 -0.2020660 5.6048453
2 ID_802 -1.192601 2.048303 -3.9976810 1.6124780
3 ID_803 3.872388 2.347984 0.8691089 6.8756664
4 ID_804 -3.218260 2.365584 -6.2327734 -0.2037462
5 ID_805 2.755695 2.486593 -0.3349594 5.8463492
6 ID_806 1.973755 2.089906 -0.8596682 4.8071788
Returns the dataframe z with following columns:
- y$IID : Target individual IDs
- y$PRS : PRS estimates of each target individual, computed using BLUE method
- y$Variance : Variance of individual PRS, computed using BLUE method
- y$Lower_Limit : Lower limit of the individual PRS confidence interval
- y$Upper_Limit : Lower limit of the individual PRS confidence interval
We recommend producing GWAS summary statistics using scaled genotypes when applying iPRSue method. If the users have access to readily-available GWAS summary statistics which are produced using unscaled genotypes, make sure to conduct necessary adjustments to those SNP effects (iPRSue_estimates_QT() or iPRSue_estimates_BT().
The adjustment can be done using effective sample size (
$beta_{adjusted} = beta / (se . \sqrt{n})$ $se_{adjusted} = 1 / \sqrt{n}$