ProbBreed employs Bayesian statistics to analyse multi-environment trials’ data, and uses its outputs to estimate the marginal, pairwise, and conditional probability of superior performance of the genotypes. The method is thoroughly described at https://doi.org/10.1007/s00122-022-04041-y.
You can install the development version of ProbBreed from GitHub with:
# install.packages("devtools")
devtools::install_github("saulo-chaves/ProbBreed")A basic workflow using the available data is:
library(ProbBreed)
mod = bayes_met(data = soy,
gen = c("Gen", "normal", "cauchy"),
env = c("Env", "normal", "cauchy"),
rept = NULL,
reg = list(c("Reg", "normal", "cauchy"), c("normal", "cauchy")),
res.het = F,
sigma.dist = c("cauchy", "cauchy"),
mu.dist = c("normal", "cauchy"),
gei.dist = c("normal", "normal"),
trait = "eBLUE",
hyperparam = "default",
iter = 100, cores = 4, chain = 4)
# You may want to increase the number of iterations, cores and chains
outs = extr_outs(data = soy,
trait = "eBLUE",
gen = "Gen",
model = mod,
effects = c('l','g','gl','m','gm'),
nenv = length(unique(soy$Env)),
res.het = FALSE,
probs = c(0.05, 0.95)
check.stan.diag = TRUE)
margs = marg_prob(data = soy,
trait = "eBLUE",
gen = "Gen",
env = "Env",
extr_outs = outs,
int = 0.2,
increase = TRUE,
save.df = FALSE,
interactive = FALSE)
conds = cond_prob(data = soy,
trait = "eBLUE",
gen = "Gen",
env = "Env",
reg = "Reg",
extr_outs = outs,
int = 0.2,
increase = TRUE,
save.df = FALSE,
interactive = FALSE)