Jordan R. Ubbens

This paper introduces LYM-1, a multi-crop yield prediction model for Canadian prairies using extensive yield, weather, and soil data, showing improved predictive performance and revealing agronomically relevant interactions.

The paper builds contrastive models to predict expression differences between ortholog pairs using a multiomic dataset from legumes.

This paper introduces GPFN, a new genomic prediction method that outperforms linear models on multiple traits and matches their performance on a structured prediction task.

This paper develops a knowledge graph and uses graph neural networks to improve gene functional annotation in plants, outperforming orthology-based methods in transferring knowledge between species.

This paper evaluates nonlinear methods for visualizing and inferring population structure, finding graph-based algorithms like t-SNE and UMAP outperform PCA while neural networks perform worse.

This paper shows deep neural networks for genomic prediction do not estimate marker effects, instead relying on genetic relatedness.