An R Package for Multitrait and Multienvironment Data with the Item‐Based Collaborative Filtering Algorithm

Abstract

The Item Based Collaborative Filtering for Multi-Trait and Multi-Environment Data (IBCF.MTME) package was developed to implement the item based collaborative filtering (IBCF) algorithm for continuous phenotypic data in the context of plant breeding where data are collected for various traits and environments. The main difference between this package and the other available packages that can implement IBCF is that this one was developed for continuous phenotypic data, which cannot be implemented in the current packages, for they can implement IBCF only for binary and ordinary phenotypes. In the following article we will show how to both install the package and use it for studying the prediction accuracy of multiple-trait and multiple-environment data under phenotypic and genomic selection. We illustrate its use with six examples (with the information of two data sets, Wheat_IBCF and Year_IBCF, that are included in the package) comprising multi-environment data, multi-trait data and both multi-trait and multi-environment data that cover scenarios in which breeding scientists are interested. The package offers many advantages for studying the genomic-enabled prediction Page 1 of 42 Plant Gen. Accepted Paper, posted 05/26/2018. doi:10.3835/plantgenome2018.02.0013 2 accuracy of multi-trait and multi-environment data, ultimately helping plant breeders make better decisions.