/metabolomicsRegression

Regression-Based Modeling of Complex Plant Traits Based on Metabolomics Data

Primary LanguageR

Regression tutorial for plant metabolomics data

This tutorial, based on the data from de Abreu e Lima et al. (2017), aims at predicting hybrid maize biomass (FW) in the field using the metabolic profiles of primary roots from seedlings (http://onlinelibrary.wiley.com/doi/10.1111/tpj.13495/full).

It covers pre-processing, cross-validation and importance analysis using OLS, PLS, LASSO, ridge, elastic net, SVM and random forests.

  • n hybrids = 362
  • n metabolites = 165

Note that the biomass data set also includes dry weight (DW), feel free to use it.