gemma2: gemma2 : GEMMA programme in R (better version of gemma)
In jumentib/useFonc: Useful Functions

Description Usage Arguments Details Value

gemma2 : GEMMA programme in R (better version of gemma)

gemma2(geno, pheno, map, covar = NULL, gemma.exe = "./gemma.macosx",
  opt.mat = c(1, 2), met.reg = c("lmm", "bslmm"), lmm = c(1, 2, 3,
  4), bslmm = c(1, 2, 3), pmin = 0.01, pmax = 0.05,
  opt.bs = "-w  10000 -s  10000", smax = 300)

`geno`	: genotype –> one row per SNP, and one column per sample.
`pheno`	: phenotype d'interet (doit etre une matrice)
`map`	: This file contains one line per SNP, with
`covar`	: covariable et NULL marche
`gemma.exe`	: the executable gemma
`opt.mat`	: option for the calcul of the related.matrix, "1" calculates the centered relatedness # matrix while "2" calculates the standardized relatedness matrix
`met.reg`	: linear mixed model of Bayesian sparse linear mixed model.
`lmm`	: if lmm choose option : where the "1" performs Wald test "2" performs likelihood ratio test, "3" performs score test, "4" performs all the three tests.
`bslmm`	: if bslmm choose option : "1" fits a standard linear BSLMM "2" fits a ridge regression/GBLUP, "3" fits a probit BSLMM.
`pmin`	: for bslmm : initial no zero proportion of beta (pi)
`pmax`	: for bslmm : max no zero proportion of beta (pi)
`opt.bs`	: option for bslmm : w and s default as in Zeng 2017 (10000 for both)
`smax`	: Number of variants with sparse effects (~ number of major effect loci). Default to 300 like in gemma
`three`	columns: (1) SNP label, (2) base-pair position, (3) chromosome.