# Any R software package that can perform GxE and stability analysis? [closed]

I am looking for a software package that can perform stability analysis as done by:

Eberhart, S.A. and W.A. Russell, 1966. Stability parameters for comparing varieties. Crop Sci., 6: 36-40.

Is a quite old method but still much appreciated. Also I would like to perform AMMI model analysis for GxE analysis.

Is there any R software package can do this ?

Check plant breeding package, on Rforge. The following is example from the package: on Stability, AMMI analysis

# stability analysis
require(plantbreeding)
data(multienv)
out <- stability (dataframe = multienv , yvar = "yield", genotypes = "genotypes",
environments = "environments", replication =  "replication")
out
# AMMI analysis
results <- ammi.full(dataframe = multienv , environment = "environments", genotype = "genotypes",
replication = "replication", yvar = "yield")


You can follow the developer on the Rforge and the blog (http://rplantbreeding.blogspot.com/)

Please, take a look at my R package metan (multi-environment trial analysis). You may perform Eberhart and Russell regression analysis with the function ge_reg(). The following code reproduces the example of the package. Note that it works naturally with the forward-pipe operator %>% and allows analyzing several traits at the same time.

# Install from Github (require devtools package)
devtools::install_github("TiagoOlivoto/metan")

library(metan)
reg <- data_ge2 %>%
ge_reg(ENV, GEN, REP,
resp = c(PH, EH))
print(reg)


AMMI analysis is performed with the function performs_ammi(). The following code reproduces the example of the package.

ammi <- data_ge2 %>%
performs_ammi(ENV, GEN, REP,
resp = c(PH, EH))

# Significance of IPCA
get_model_data(ammi, "ipca_pval")

# Explained sum of squares
get_model_data(ammi, "ipca_expl")

# Predict the response variables (say using 2 IPCA)
pred <- predict(ammi, naxis = c(2, 2))
pred\$GY

# Biplots (variable GY)
# AMMI1
plot_scores(ammi)

# AMMI2
plot_scores(ammi, type = 2)



You may find the complete description of the package at https://tiagoolivoto.github.io/metan/index.html

• You should disclose that you are the author of this package (I assume). – OrangeDog Oct 18 '19 at 13:39