Search

Sorghum (Sorghum bicolor) has the ability to inhibit the conversion of ammonium to nitrate (biological nitrification inhibition [BNI]) in the rhizosphere, which in turn prevents the loss of bioavailable nitrogen. Sorgoleone is a lipidic compound secreted by sorghum root hairs and is responsible for roughly 60% of BNI activity in sorghum. Previous studies revealed variation in sorgoleone secretion among different accessions and cultivars. However, little information is available regarding the genetic inheritance of sorgoleone secretion in sorghum. To increase sorgoleone through breeding, an understanding of the inheritance of this trait is required. In this study, 21 seed parents and 21 pollinator parents from the Texas AgriLife Research sorghum breeding program were crossed in an incomplete factorial design to generate 158 hybrids, and sorgoleone secretion from both hybrids and inbreds was quantified. There was significant variation in sorgoleone secretion across hybrids and inbred lines, and small but significant mid‐parent heterosis was observed in the hybrids. A linear mixed model analysis to calculate general and specific combining abilities for inbred parents and hybrids detected significant genetic effects for the male, the female, and male × female interactions (p < 0.001). Broad‐sense heritability was high (H2${{H}^2}$ = 0.87), while narrow‐sense heritability for the seed parents and pollinator parents was moderate (Hf2$H_f^2$= 0.35 and Hm2$H_m^2$ = 0.39, respectively). These results indicate that sorgoleone exudation is primarily driven by additive genetic effects, but dominance effects are important for optimum production. These findings indicate that selection for increased sorgoleone root exudation among elite, adapted grain sorghum hybrids and inbred should be effective. [ABSTRACT FROM AUTHOR]

Sorghum (Sorghum bicolor [L. Moench]) is among the most important cereal crops worldwide and is widely grown across Texas’ diverse geographies. While sorghum breeding programs strive to improve both profitability and adaptation of the crop, increases in grain yield are the ultimate measure of progress. Using five decades of data compiled by Texas A&M AgriLife Research, trends in grain sorghum were retrospectively assessed. Performance data from over 4094 hybrids were collected between 1970 and 2021 in a total of 514 county‐year environments. Hybrids in these trials were either commercially available or in the final stages of testing for release. A linear mixed model approach produced the best linear unbiased predictions of genotypic and environmental effects for grain yield and relevant agronomic traits. These measures, as well as the raw data, were used to inform and visualize trait dependencies, yield gains, and geographic regions with similar effects on hybrid performance. Over these decades, grain yield increased at an average rate of 0.03 t ha−1 per year. A strong positive relationship between hybrid yield and plant height, as well as several other correlations, were detected through separate periods of yield increases which occurred in the early 1980s and 1990s; since that time, a sinusoidal pattern of yield performance has prevailed. Principal component analysis clustered production regions primarily based on moisture availability. The results herein indicate that modest increases in height can increase grain yield and that available moisture is the preeminent delineator of grain sorghum production environments in Texas. [ABSTRACT FROM AUTHOR]

Gleason grade stratification is the main histological standard to determine the severity and progression of prostate cancer. Nonetheless, there is a high variability on disease diagnosis among expert pathologists (kappa lower than 0.44). End-to-end deep representations have recently deal with the automatic classification of Gleason grades, where each grade is limited to namely code high-visual-variability sharing patterns among classes. Such limitation on models may be attributed to the relatively few labels to train the representation, as well as, to the natural imbalanced sets, available in clinical scenarios. To overcome such limitation, this work introduces a new embedding representation that learns intra and inter-Gleason relationships from more challenging class samples (grades tree and fourth). The proposed strategy implements a triplet loss scheme building a hidden embedding space that correctly differentiates close Gleason levels. The proposed approach shows promising results achieving an average accuracy of 74% to differentiate between degrees three and four. For classification of all degrees, the proposed approach achieves an average accuracy of 62%.