HPQTL is R package for quantitative trait locus (QTL) analysis in case of highly polygenic genetic backgound.
To calculate LOD scores, the user need to specify the following
- phenotype (variable of interest + covariates)
- genotype (3-dim array of probabilities: subjects x calls x markers)
- genetic similarity matrix (can be calculated from genotype)
Three methods have been implemented
- linear model (LM)
- linear mixed model (LMM)
- linear mixed model with specific genetic similarity matrix for each chromosome (LMM-L1O)
library(HPQTL)
data(fake.f2, package="qtl")
# calculate and extract genotype probabilities
fake.f2 <- qtl::calc.genoprob(fake.f2)
geno <- extract.geno(fake.f2)
# calculate genetic similarity matrix
G <- gensim.matrix(geno)
# mapping with linear model
qtl::scanone(fake.f2, method = "hk")
(fit.lm <- scan1(geno=geno, fake.f2$pheno))
# mapping with linear mixed model
fit.lmm <- scan1(geno=geno, fake.f2$pheno, procedure = "LMM", G=G)
# mapping with linear mixed model - leave the scanned chromosome out
fit.lmm_loco <- scan1(geno=geno, fake.f2$pheno, procedure = "LOCO")
# LOD plots
plot(fit.lm, col="black", incl.markers=FALSE)
plot(fit.lmm, add=TRUE, col="red")
plot(fit.lmm_loco, add=TRUE, col="blue")
legend("topleft", c("LM", "LMM", "LOCO"), lty=1, col=c("black", "red", "blue"))
To install the package directly from Github, use devtools::install_github:
library(devtools)
install_github("simecek/HPQTL")
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent