README

ProbBreed employs Bayesian statistics to analyse multi-environment trials’ data, and uses its outputs to estimate the marginal and pairwise probabilities of superior performance and superior stability of the genotypes, as well as their conditional probability of superior performance. The method is thoroughly described at https://doi.org/10.1007/s00122-022-04041-y.

Installation

install.packages("ProbBreed")

Alternatively, you can install the development version of ProbBreed from GitHub with:

# install.packages("devtools")
devtools::install_github("saulo-chaves/ProbBreed")

Usage

library(ProbBreed)

Currently, ProbBreed has nine available models implemented in the bayes_met function. See ?bayes_met for more details. Using the soy example dataset, a basic usage is as follows:

model = bayes_met(data = soy,
                gen = "Gen",
                loc = "Loc",
                repl = NULL,
                year = NULL,
                reg = NULL,
                res.het = TRUE,
                trait = 'Y',
                iter = 6000, cores = 4, chains = 4)

gen, loc, repl, year and reg are all column names that contain information on genotypes, environments (locations), replicates, years (or seasons) and regions (or mega-environments). The soy dataset has only adjusted means, so only gen and loc have values. res.het indicates if a per-environmental residual variance should be estimated. trait is the column in data that contain the phenotypic observations. The other arguments are specifications for model fitting: the number of iterations, cores and chains. Feel free to customize these and other options according to your necessity.

The output of this function will be an object of class stanfit, which should be used in the extr_outs function for further processing before computing the probabilities per se. This function also provides some useful diagnostics. Here is how to use it:

outs = extr_outs(model = model,
                 probs = c(0.05, 0.95),
                 verbose = TRUE)

The object of class extr provided by this function contains the effects’ posterior and maximum posterior, the models’ variance components and some posterior predictive checks. Here are them:

outs$variances
#>         effect     var      sd naive.se HPD_0.05 HPD_0.95
#> 1          Gen   3.383   1.283    0.012    1.655    5.758
#> 2          Loc 246.354 124.591    1.137  115.921  470.127
#> 3   error_env1  10.077   2.679    0.024    6.418   15.058
#> 4   error_env2  28.924   7.296    0.067   19.029   42.483
#> 5   error_env3  11.620   3.372    0.031    7.210   17.970
#> 6   error_env4  18.458   4.732    0.043   12.092   27.309
#> 7   error_env5  51.193  12.312    0.112   34.576   73.887
#> 8   error_env6  15.524   4.085    0.037   10.051   23.073
#> 9   error_env7  19.385   5.003    0.046   12.723   28.555
#> 10  error_env8  21.217   5.390    0.049   13.971   31.041
#> 11  error_env9  14.818   3.862    0.035    9.658   21.890
#> 12 error_env10  13.633   5.612    0.051    7.024   24.165
#> 13 error_env11  22.330   9.116    0.083   11.656   39.590
#> 14 error_env12   7.657   3.377    0.031    3.693   14.020
#> 15 error_env13  14.604   5.965    0.054    7.650   25.520
#> 16 error_env14  10.983   4.557    0.042    5.678   19.356
outs$ppcheck
#>                   Diagnostics
#> p.val_max              0.9218
#> p.val_min              0.3498
#> p.val_median           0.7276
#> p.val_mean             0.5012
#> p.val_sd               0.5630
#> Eff_No_parameters     27.4466
#> WAIC2               2715.4466
#> mean_Rhat              1.0002
#> Eff_sample_size        0.8112

You can also the plot S3 method for some useful visualizations. For e.g., the comparison between the empirical and sampled phenotype illustrates the model’s convergence:

After these two steps, everything is set to compute the probabilities. This can be done using the function prob_sup:

results = prob_sup(extr = outs, 
                   int = .2,
                   increase = TRUE, 
                   save.df = FALSE, 
                   verbose = TRUE)

This function generates an object of class probsup, which contains two lists: across and within. As their names suggest, the across list has the across-environments probabilities, and is suitable for a broader recommendation. Conversely, the within results are more appropriate to specific recommendations. For example, here are some probability of superior performances across and within environments:

head(results$across$perfo)
#>     ID      prob
#> 36 G36 0.9845833
#> 9  G09 0.8344167
#> 20 G20 0.8135833
#> 38 G38 0.7305833
#> 31 G31 0.6705000
#> 1  G01 0.5160000
head(results$within$perfo$gl)
#>   gen          E01          E02         E03          E04          E05
#> 1 G01 0.5160000000 0.5160000000 0.656916667 0.5160000000 0.5160000000
#> 2 G02 0.0062500000 0.0062500000 0.013916667 0.0062500000 0.0062500000
#> 3 G03 0.1638333333 0.1638333333 0.247416667 0.1638333333 0.1638333333
#> 4 G04 0.0138333333 0.0138333333 0.029250000 0.0138333333 0.0138333333
#> 5 G05 0.0008333333 0.0008333333 0.002583333 0.0008333333 0.0008333333
#> 6 G06 0.1968333333 0.1968333333 0.290083333 0.1968333333 0.1968333333
#>            E06          E07          E08          E09       E10       E11
#> 1 0.5160000000 0.5160000000 0.5160000000 0.5160000000 0.9541667 0.9541667
#> 2 0.0062500000 0.0062500000 0.0062500000 0.0062500000        NA        NA
#> 3 0.1638333333 0.1638333333 0.1638333333 0.1638333333        NA        NA
#> 4 0.0138333333 0.0138333333 0.0138333333 0.0138333333        NA        NA
#> 5 0.0008333333 0.0008333333 0.0008333333 0.0008333333        NA        NA
#> 6 0.1968333333 0.1968333333 0.1968333333 0.1968333333        NA        NA
#>         E12       E13       E14
#> 1 0.9541667 0.9541667 0.9541667
#> 2        NA        NA        NA
#> 3        NA        NA        NA
#> 4        NA        NA        NA
#> 5        NA        NA        NA
#> 6        NA        NA        NA

The S3 method plot is also available for probsup objects. Here are some of them:

plot(results, category = "pair_perfo", level = "across")

plot(results, category = "stabi")

plot(results, category = "perfo", level = "within")

The grey cells are environments where a given genotype was not assessed. See more options at ?plot.probsup.

The estimation of these probabilities are strictly related to some key questions that constantly arises in plant breeding, like:

ProbBreed

Installation

Usage

Citation