SNP data and phenotypes for oil palm training and test sets, and associated script for genomic predictions with multi-layer perceptron

The data come from two experimental designs located on the SOCFINDO estate in North Sumatra, Indonesia. One is used to train genomic prediction models, the other as test set. The SNP genotypes were obtained with a custom Affymetrix Axiom® SNP array. The phenotypes are fresh fruit bunch production (FFB) corresponding to annual cumulative bunch (kg), bunch number (BN) corresponding to annual cumulative bunch number, and annual average bunch weight (BW, in kg). The python script implements genomic prediction with the multi-layer perceptron artificial neural network.