Dariva, F. | Search bioRxiv

bioRxiv2024

High-Throughput Phenotyping of Seed Quality Traits Using Imaging and Deep Learning in Dry Pea

Seed traits, such as seed color and seed size, directly impact seed quality, affecting the marketability and value of dry peas [1]. Assessing seed quality is integral to a plant breeding programs to ensure optimal seed standards. This research introduced a phenotyping tool to assess seed quality traits specifically tailored for pulse crops, which integrates image processing with cutting-edge deep learning models. The proposed method is designed for automation, seamlessly processing a sequence of images while minimizing human intervention. The pipeline standardized red-green-blue (RGB) images captured from a color light box and used deep learning models to segment and detect seed features. Our method extracted up to 86 distinct seed characteristics, ranging from basic size metrics to intricate texture details and color nuances. Compared to traditional methods, our pipeline demonstrated a 95 percent similarity in seed quality assessment and increased time efficiency (from 2 weeks to 30 minutes for processing time). Specifically, we observed an improvement in the accuracy of seed trait identification by simply using an RGB value instead of a categorical, non-standard description, which allowed for an increase in the range of detectable seed quality characteristics. By integrating conventional image processing techniques with foundational deep learning models, this approach emerges as a pivotal instrument in pulse breeding programs, guaranteeing the maintenance of superior seed quality standards.

plant biology↗

bioRxiv2024

Multi-trait multi-environment genomic prediction of preliminary yield trials in pulse crops

Phenotypic selection in preliminary yield trials (PYT) is challenged by limited seeds, resulting in trials with few replications and environments. The emergence of multi-trait multi-environment enabled genomic prediction (MTME-GP) offers opportunity for enhancing prediction accuracy and genetic gain across multiple traits and diverse environments. Using a set of 300 advanced breeding lines in the North Dakota State University (NDSU) pulse crop breeding program, we assessed the efficiency of a MTME-GP model for improving seed yield and protein content in field peas in stress and non-stress environments. MTME-GP significantly improved predictive ability, improving up to 2.5-fold, particularly when a significant number of genotypes overlapped across environments. Heritability of the training environments contributed significantly to the overall prediction of the model. Average predictive ability ranged from 3 to 7-folds when environments with low heritability were excluded from the training set. Overall, the Reproducing Kernel Hilbert Spaces (RKHS) model consistently resulted in improved predictive ability across all breeding scenarios considered in our study. Our results lay the groundwork for further exploration, including integration of diverse traits, incorporation of deep learning techniques, and the utilization of multi-omics data in predictive modeling. Core ideasO_LIPhenotypic selection in PYT is challenged by limited seeds, resulting to few replications and environments. C_LIO_LIMTME-GP offers opportunity for enhancing prediction accuracy of multi-trait and diverse environments in PYT. C_LIO_LIMTME-GP enhances prediction by up to 2.5-fold, especially with numerous overlapping genotypes in various tested environments. C_LIO_LIRKHS MTME-GP models, excels in low-heritability, negatively correlated traits, like drought-affected conditions. C_LI

genomics↗