The study introduces a deep‑learning based image analysis pipeline that scores blackleg disease severity from stem cross‑section images of canola species, achieving greater consistency than median expert raters while preserving comparable heritability of susceptibility traits. This standardized scoring method aims to improve selection of resistant varieties in breeding programs.

The study presents a deep‑learning pipeline that uses state‑of‑the‑art convolutional neural networks to automatically estimate the establishment of perennial groundcovers in agricultural research plots from smartphone images. By employing region‑of‑interest markers and deploying the models on AWS SageMaker with a lightweight Django web interface, the approach provides fast, objective, and reproducible assessments that can be adopted by researchers and growers across the Midwest.

The study applied the STOmics spatial transcriptomics platform to map gene expression at subcellular resolution in developing wheat (Triticum aestivum) seeds during grain filling, analyzing over four million transcripts. Eight functional cellular groups were identified, including four distinct endosperm clusters with radial expression patterns and novel marker genes, and subgenome‑biased expression was observed among specific paralogs. These results highlight spatial transcriptomics as a powerful tool for uncovering tissue‑specific and polyploid‑specific gene regulation in seeds.

The study combined spatial transcriptomics and single-nuclei RNA sequencing to map soybean (Glycine max) responses to Asian soybean rust caused by Phakopsora pachyrhizi, revealing two distinct host cell states: pathogen‑occupied regions and adjacent non‑infected regions that show heightened defense gene expression. Gene co‑expression network analysis identified a key immune‑related module active in the stressed cells, highlighting a cell‑non‑autonomous defense mechanism.

The study used an interpretable convolutional neural network to predict ABA responsiveness from proximal promoter sequences in Arabidopsis thaliana, revealing both known ABF-binding motifs and novel regulatory elements. Model performance was boosted by advanced data augmentation, and predicted regulatory regions were experimentally validated using reporter lines, confirming the inferred cis‑regulatory code.

The study uses an imputation strategy that integrates deep single-cell RNA sequencing with spatial gene expression data to map transcriptional dynamics across barley inflorescence development at cellular resolution. By leveraging the BARVISTA web interface, the authors identify key transcriptional events in meristem founder cells, characterize complex branching mutants, and reconstruct spatio‑temporal trajectories of flower organogenesis, offering insights for targeted trait manipulation.

The study presents GenoRetriever, an interpretable deep learning framework trained on STRIPE-seq data from soybean and other crops, that predicts transcription start site locations and usage by identifying 27 core promoter motifs. Validation using in silico motif insertions, saturation mutagenesis, and CRISPR‑Cas9 promoter editing demonstrates high predictive accuracy and reveals domestication‑driven motif usage shifts and lineage‑specific effects. The tool is provided via a web server for promoter analysis and design, offering a new resource for plant functional genomics and crop improvement.

The Global Wheat Dataset Consortium released a comprehensive semantic segmentation dataset (GWFSS) of wheat organs across developmental stages, comprising 1,096 fully annotated images and 52,078 unannotated images from 11 institutions. Models based on DeepLabV3Plus and Segformer were trained, with Segformer achieving ≈90% mIoU for leaves and spikes but lower precision (54%) for stems, while also enabling weed exclusion and discrimination of necrotic, senescent, and residue tissues.

The study introduces an in-soil fiber Bragg grating (FBG) sensing system that continuously records three-dimensional strain from growing pseudo-roots, enabling non‑destructive monitoring of root architecture. Using two ResNet models, the system predicts root width and depth with over 90% accuracy, and performance improves to 96‑98% after retraining on data from actual corn (Zea mays) roots over a 30‑day period. This prototype demonstrates potential for scalable, real‑time root phenotyping and broader soil environment sensing.