The study introduces a hybrid modeling framework that integrates a logistic ordinary differential equation with a Long Short-Term Memory neural network to form a Physics-Informed Neural Network (PINN) for predicting wheat plant height. Using only time and temperature as inputs, the PINN outperformed other longitudinal growth models, achieving the lowest average RMSE and reduced variability across multiple random initializations. The results suggest that embedding biological growth constraints within data‑driven models can substantially improve prediction accuracy for plant traits.

The study characterizes the plant spliceosomal protein AtSF1 in Arabidopsis thaliana, using iCLIP and RNA‑seq to map its in vivo branch point binding sites and demonstrate that loss of AtSF1 causes widespread 3' splice‑site mis‑selection. Structural comparison reveals a plant‑specific domain architecture, and the identified AtSF1 targets are enriched for circadian and defense genes, linking splicing regulation to timing and immunity.

The study reports the creation and validation of a high‑density Axiom SNP array for Solanum tuberosum, based on 10X Genomics sequencing of 108 diverse clones and integration of existing Illumina markers. The array demonstrated high reproducibility and, after filtering, provided 206,616 informative markers for population structure analysis, GWAS of polyphenol oxidase activity, and genomic prediction with accuracies up to 0.86.