Philosophy of phenomic prediction and its incompatibility with causal inference
Open Access
- 29 June 2026
- journal article
- Published by Wiley in The Plant Phenome Journal
- Vol. 9 (1)
Abstract
Breeding programs need to make decisions frequently to improve populations and develop varieties efficiently. These needs led to the development of genomic prediction in the early 2000s and phenomic prediction in the mid‐2010s. In practice, phenomic prediction techniques rely on the same statistical tools and computational frameworks as other analyses (e.g., genomic prediction, variance component estimation, and heritability analysis); beyond that, they share no further similarities. Phenomic prediction should excel at predicting phenotypic values, whereas genomic prediction should excel at predicting breeding values, or total genetic values if nonadditive variance is incorporated. Phenomic prediction, as commonly implemented, should not be interpreted within a causal inference framework. This is because phenomic features often share a causal structure with the focal trait (e.g., genotype and environment) and may also be statistically related to the trait itself. Consequently, including phenomic features as predictors alongside shared determinants in linear (mixed) models can induce confounding and/or collider bias. For this reason, estimated effects in phenomic prediction models should be interpreted as associations that support prediction, rather than as evidence of causal relationships. We discuss these issues and propose the multivariate best linear unbiased predictor model as a solution for phenomic prediction within a framework more amenable to causal interpretation.Keywords
Funding Information
- California Strawberry Commission
- Department of Plant Sciences, University of California, Davis
- National Institute of Food and Agriculture (2022‐51181‐38328)
- National Institute of Food and Agriculture (2024‐70412‐43648)
- National Institute of Food and Agriculture (026418H)
- National Institute of Food and Agriculture (2024‐67013‐42591)
This publication has 100 references indexed in Scilit:
- Unoccupied aerial systems imagery for phenotyping in cotton, maize, soybean, and wheat breedingCrop Science, 2023
- An adjusted coefficient of determination (R2) for generalized linear mixed models in one goBiometrical Journal, 2023
- Pedigree‐management‐flight interaction for temporal phenotype analysis and temporal phenomic predictionThe Plant Phenome Journal, 2023
- Collider BiasJAMA, 2022
- High throughput can produce better decisions than high accuracy when phenotyping plant populationsCrop Science, 2021
- Phenomic selection is competitive with genomic selection for breeding of complex traitsThe Plant Phenome Journal, 2021
- Images carried before the fire: The power, promise, and responsibility of latent phenotyping in plantsThe Plant Phenome Journal, 2021
- Phenomic selection and prediction of maize grain yield from near‐infrared reflectance spectroscopy of kernelsThe Plant Phenome Journal, 2020
- A coefficient of determination (R2) for generalized linear mixed modelsBiometrical Journal, 2019
- A stage‐wise approach for the analysis of multi‐environment trialsBiometrical Journal, 2012