---

GWASbyGLM

The goal of GWASbyGLM is to perform Genome Wide Association Studies using Generelized linear model. The final goal is to assess the association of genotypes with traits.

Installation

You can install the development version from GitHub with:

# installation of package magick is required before installing this package
install.packages("magick")
# install.packages("devtools")
devtools::install_github("atitparajuli2020/GWASbyGLM")

Example

This is a basic example which shows you how to solve a common problem:

library(GWASbyGLM)
## basic example code

Import Data

#Import Data
myGD=import_data(filename = "http://zzlab.net/GAPIT/data/mdp_numeric.txt")
myGM=import_data(filename = "http://zzlab.net/GAPIT/data/mdp_SNP_information.txt")
phenotype=import_data(filename="http://zzlab.net/GAPIT/data/CROP545_Phenotype.txt")

Perform PCA

#This functions perform Principal Component across the Markers.
# Exclude PCs linearly dependent with covariates (if there is any covariate information available)
cofactors_PC= PC(2,myGD[,-1],C=NULL);head(cofactors_PC) 
#> Warning in rm(c): object 'c' not found
#> Warning in rm(tmp): object 'tmp' not found
#>              PC1        PC2
#> [1,]   1.6780775 -4.9373382
#> [2,]  -1.6021749 -4.7322790
#> [3,]  -0.8999517 -6.2186090
#> [4,]   2.1334477 -6.2879301
#> [5,]   0.6302372 -4.8947416
#> [6,] -13.6690754  0.8736302

Perform GWAS

#This function returns the p-values for each Markers
p_val=p_val_GLM(y=phenotype[,-1],X=myGD[,-1],PC=cofactors_PC)
#Manhattan Plot
manhattan_plot(data=myGM,cutoff = 0.05,p=p_val)

#QQPlot
QQPlot_GWAS(p=p_val)

#False Positive
source("http://zzlab.net/StaGen/2020/R/G2P.R")
X=myGD[,-1]
index1to5=myGM[,2]<6
X1to5 = X[,index1to5] # Subset genotypic matrix based on Chromosome 1-5
mySim=G2P(X=X1to5,h2=0.75,alpha=1,NQTN=10,distribution="normal")
y=mySim$y
fp_GLM=false_positive(P=p_val,QTN.position = mySim$QTN.position,cutoff = 0.05);fp_GLM
#> [1] 0.03491756