| Type: | Package |
| Title: | Score Test Integrated with Empirical Bayes for Association Study |
| Version: | 0.1.1 |
| Author: | Wenlong Ren |
| Maintainer: | Wenlong Ren <wenlongren@ntu.edu.cn> |
| Description: | Perform association test within linear mixed model framework using score test integrated with Empirical Bayes for genome-wide association study. Firstly, score test was conducted for each marker under linear mixed model framework, taking into account the genetic relatedness and population structure. And then all the potentially associated markers were selected with a less stringent criterion. Finally, all the selected markers were placed into a multi-locus model to identify the true quantitative trait nucleotide. |
| License: | GPL-3 |
| Imports: | data.table |
| Encoding: | UTF-8 |
| LazyData: | true |
| NeedsCompilation: | yes |
| Depends: | R (≥ 3.5.0) |
| RoxygenNote: | 7.1.1 |
| Packaged: | 2021-09-15 15:40:45 UTC; ThinkPad |
| Repository: | CRAN |
| Date/Publication: | 2021-09-15 21:10:12 UTC |
Preconditioned Conjugate Gradient
Description
Conduct preconditioned conjugate gradient method to accelerate.
Usage
PCG(G,b,m.marker,sigma.k2,sigma.e2,tol,miter)
Arguments
G |
genotype data. |
b |
column vector. |
m.marker |
the number of markers. |
sigma.k2 |
variance of polygenic. |
sigma.e2 |
variance of residual error. |
tol |
convergence threshold. |
miter |
the maximum number of iterations. |
Value
x |
x is approximate solution of linear equations. |
Examples
data(geno)
G <- t(geno[,-c(1:4)])
n.sample <- dim(G)[1]
m.marker <- dim(G)[2]
b <- rnorm(n.sample)
sigma.k2 <- 6.0
sigma.e2 <- 10.0
tol <- 5e-4
miter <- 20
PCG(G,b,m.marker,sigma.k2,sigma.e2,tol,miter)
Score Test Integrated with Empirical Bayes for Association Study
Description
Perform association test within linear mixed model framework using score test integrated with Empirical Bayes for genome-wide association study. Firstly, score test was conducted for each marker under linear mixed model framework, taking into account the genetic relatedness and population structure. And then all the potentially associated markers were selected with a less stringent criterion. Finally, all the selected markers were placed into a multi-locus model to identify the true quantitative trait nucleotide.
Usage
ScoreEB(genofile, phenofile, popfile = NULL, trait.num = 1, EMB.tau = 0,
EMB.omega = 0, B.Moment = 20, tol.pcg = 1e-4, iter.pcg = 100, bin = 100,
lod.cutoff = 3.0, seed.num = 10000, dir_out)
Arguments
genofile |
Genotype file name, change the file path where it is located, i.e.,"D:/Genotype_Example.csv". |
phenofile |
Phenotype file name, change the file path where it is located, i.e.,"D:/Phenotype_Example.csv". |
popfile |
Population structure file name, change the file path where it is located,i.e.,"D:/Population.csv". |
trait.num |
trait.num stands for computing trait from the 1st to the "trait.num" |
EMB.tau |
EMB.tau and EMB.omega are two values of hyperparameters in empirical Bayes step, which are set to 0 by default. |
EMB.omega |
As describe in EMB.tau |
B.Moment |
B.Moment is a parameter to obtain trace of NxN matrix approximately using method of moment. B.Moment is set to 20 by default. |
tol.pcg |
tol.pcg and iter.pcg are tolerance and maximum iteration number in preconditioned conjugate gradient algorithm. |
iter.pcg |
As describe in tol.pcg |
bin |
bin is to choose the maximum score within a certain range. |
lod.cutoff |
lod.cutoff is the threshold to determine identified QTNs. |
seed.num |
Set a random number. |
dir_out |
Give the path where it will be saved,i.e.,"D:/Result" |
Value
result.total |
A data frame of identified markers, including "Trait", "Id", "Chr", "Pos", "Score", "Beta", "Lod" and "Pvalue" of markers. |
Note
1. genofile and phenofile are the required input file, while popfile is the optional input file.
2. In the "tempdir()" folder, there are two results files "ScoreEB.Result.csv" and "ScoreEB.time.csv" generated and saved after the run.
3. The results file "ScoreEB.Result.csv" has 8 columns, including "Trait", "Id", "Chr", "Pos", "Score", "Beta", "Lod" and "Pvalue".
4. The time file "ScoreEB.time.csv" includes 3 rows, which are "User", "System", "Elapse" time, respectively.
Author(s)
Wenlong Ren
Wenlong Ren <wenlongren@ntu.edu.cn>
Examples
genofile <- system.file("extdata", "Genotype_Example.csv", package="ScoreEB")
phenofile <-system.file("extdata", "Phenotype_Example.csv", package="ScoreEB")
dir_out <- tempdir()
ScoreEB(genofile, phenofile, popfile = NULL, trait.num = 1, EMB.tau = 0,
EMB.omega = 0, B.Moment = 20, tol.pcg = 1e-4, iter.pcg = 100, bin = 100,
lod.cutoff = 3.0, seed.num = 10000, dir_out)
Empirical Bayes for multi-locus selection
Description
Empirical Bayes using expectation–maximization algorithm.
Usage
ebayes_EM(x,z,y,EMB.tau,EMB.omega)
Arguments
x |
fixed effect vector or matrix. |
z |
genotype data. |
y |
phenotype data. |
EMB.tau |
one of hyperparameters in inverse chi-square distribution. |
EMB.omega |
one of hyperparameters in inverse chi-square distribution. |
Value
u |
The effect values of markers, and their absolute values are used as the basis for further screening. |
Examples
data(geno)
data(pheno)
EMB.tau <- 0
EMB.omega <- 0
z <- t(geno[,-c(1:4)])
y <- as.matrix(pheno)
nsample <- dim(z)[1]
x <- as.matrix(rep(1,nsample))
ebayes_EM(x,z,y,EMB.tau,EMB.omega)
Genotype of example data
Description
Genotype dataset with SNP chromosome, position and etc.
Usage
data(geno)
Details
Dataset input of genotype in ScoreEB function.
Carry out likelihood ratio test
Description
Snps selected via EM-Bayes to further identified by likelihood ratio test.
Usage
likelihood(xxn,xxx,yn,bbo)
Arguments
xxn |
fixed effect vector or matrix. |
xxx |
snp matrix which are selected by EM-Bayes. |
yn |
phenotype data. |
bbo |
effect value of snp estimated by EM-Bayes. |
Value
lod |
Odds of logarithm vector of markers. |
Examples
data(geno)
data(pheno)
z <- t(geno[,-c(1:4)])
y <- as.matrix(pheno)
n.sample <- dim(z)[1]
m.marker <- dim(z)[2]
x <- as.matrix(rep(1,n.sample))
beta <- rnorm(m.marker)
likelihood(x,z,y,beta)
Multivariate normal distribution
Description
Obtain P value with multivariate normal distribution.
Usage
multinormal(y,mean,sigma)
Arguments
y |
column vector. |
mean |
arithmetic mean. |
sigma |
standard deviation. |
Value
pdf_value |
A vector of multivariate normal distribution density function. |
Examples
data(pheno)
y <- pheno
mean <- 2.0
sigma <- 1.5
multinormal(y,mean,sigma)
Phenotype of example data
Description
Phenotype dataset of multiple traits.
Usage
data(pheno)
Details
Dataset input of phenotype in ScoreEB function.