README.md

In JunhuiLi1017/JM4QTN: Joint Mapping for QTN

JM4QTN:Joint Map for QTN

1. Introduction

• What is the package JM4QTN?

** Joint Mapping for QTN, it combines composite interval mapping(CIM) and regression analysis to do association mapping and linkage mapping and identify pleiotropic or linked QTL in multiple-traits both in single or multiple line cross populations. It can also compute conditional probability of missing marker genotype using flanking markers. Thresholds are determined by permutation test and parametric bootstrap test. Stepwise regression with multivariate and univariate is carried out with several information criteria to select cofactors across families.

** The main module in this package

(1) Best Linear Unbiased Eestimation of Phenotype

(2) Missing marker genotype estimation based on different mapping population(RIL, F2, Fn, BCP1, BCP2, DH)

(3) Threshold of LOD and p value calculation

(4) QTL scan for multiple-line cross population

(5) Pleiotropic and linked QTL detection

2. Statistical and coding details in this package

• What is the main features in JM4QTN?

** Missing marker genotype estimation: Hidden Markov model based on genotype frequency for different mapping population(RIL, F2, Fn, BCP1, BCP2, DH)

** Linear model selection (stepwise regression) programmed with Cpp: (1)Variable enter direction: forward stepwise selection and hybrid approaches (2)Information criteria: AIC/AICc/BIC/SBC/HQ and Significant Level (3)Multicollinearity: tolerance and VIF (4)Independent variables: continuous effects and continuous-nesting-class effects (5)Dependent variables: univariate and multivariate

** P/LOD value calculation(likelihood ratio test): (1)F test for univariate based on full and reduced linear models (2)Approximate F test for multivariate based on Wilks' lambda, Pillai’s' Trace and Hotelling-Lawley Trace with full and reduced linear models

** Resample:(1)Permutation test (2)bootstrap and (3)cross-validation

3. Usage and Examples

#install.package("JM4QTN")

library(JM4QTN)

#load data
data(GeneticMap)
data(GenoData)
data(PhenoData)
data(GenoData_EST)
data(GenoData_S2)

#Estimate missing markers: step=0 and calculate conditional probability

cross <- "RIL"
method <- "LM"    #"AM"
steps <- 0
vecPheno <- c("newEC1","newEC2")
Data_calGenoProb <- calGenoProb(GeneticMap,MarkerData,method,cross,0)
str(Data_calGenoProb);dim(Data_calGenoProb);Data_calGenoProb[1:10,1:10]

# Calculate QTL conditional probability with max step=2

steps <- 2
RMdata_2 <- calGenoProb(GeneticMap,MarkerData,method,cross,steps)
RMdata_2[1:10,1:20];dim(RMdata_2)


#permutation test
vecPheno <- c("newEC1","newEC2")
#vecPheno <- c("newEC1")
croType <- "RIL"
GenoData_QTL <- GenoData_S2
vecThrVal <- ptJM1(vecPheno,PhenoData,GenoData_EST,method,npt=100,alpha=0.1,selection="stepwise",tolerance=1e-7,Trace="Pillai",select="SBC",sle=0.05,sls=0.05,Choose="SBC")
#vecThrVal <- c(0.0001,0.0001,3.3,3.2)

# QTL scan
vecH2 <- c(0.8,0.8)
JM(vecPheno,vecH2,PhenoData,method,vecThrVal,GenoData_EST,GenoData_QTL)

# Pleiotropic QTL detection
vecThrVal <- ptLP1(vecPheno,PhenoData,GenoData_EST,npt=100,alpha=0.1,selection="stepwise",tolerance=1e-7,Trace="Pillai",select="SBC",sle=0.05,sls=0.05,Choose="SBC")
CChr <- 2
#vecThrVal <- c(0.000206,4.62,4.70,6.64)
Interval <- c(56.896,57.237,57.237,61.323)
LP(vecPheno,PhenoData,GenoData_EST,GenoData_QTL,vecThrVal,CChr,Interval,nPB=200,alpha=0.05)

JunhuiLi1017/JM4QTN documentation built on June 4, 2019, 4:10 a.m.