Heterosis in locally adapted sorghum genotypes and potential of hybrids for increased productivity in contrasting environments in Ethiopia

Increased productivity in sorghum has been achieved in the developed world using hybrids.Despite their yield advantage,introduced hybrids have not been adopted in Ethiopia due to the lack of adaptive traits,their short plant stature and small grain size.This study was conducted to investigate hybrid performance and the magnitude of heterosis of locally adapted genotypes in addition to introduced hybrids in three contrasting environments in Ethiopia.In total,139 hybrids,derived from introduced seed parents crossed with locally adapted genotypes and introduced R lines,were evaluated.Overall,the hybrids matured earlier than the adapted parents,but had higher grain yield,plant height,grain number and grain weight in all environments.The lowland adapted hybrids displayed a mean better parent heterosis(BPH) of19%,equating to 1160 kg ha-1and a 29% mean increase in grain yield,in addition to increased plant height and grain weight,in comparison to the hybrids derived from the introduced R lines.The mean BPH for grain yield for the highland adapted hybrids was 16% in the highland and 52%in the intermediate environment equating to 698 kg ha-1and 2031 kg ha-1,respectively,in addition to increased grain weight.The magnitude of heterosis observed for each hybrid group was related to the genetic distance between the parental lines.The majority of hybrids also showed superiority over the standard check varieties.In general,hybrids from locally adapted genotypes were superior in grain yield,plant height and grain weight compared to the high parents and introduced hybrids indicating the potential for hybrids to increase productivity while addressing farmers' required traits.

A Graph-Based Pan-Genome Guides Biological Discovery

The availability of reference genomes has been the foundation for genomics research in the last decade.However,as reports of presence/absenee variations of sequence and genes are growing in various species(Zhang et al.,2016;Sun et al.,,2018)it is increasingly evident that a single reference genome provides an in adequate represe ntation of the whole landscape of sequence diversity within a species.Pan-genome analysis provides a platform to investigate the entire genome repertoire of a biological clade.

A Weakly Supervised Deep Learning Framework for Sorghum Head Detection and Counting

The yield of cereal crops such as sorghum(Sorghum bicolor L.Moench)depends on the distribution of crop-heads in varying branching arrangements.Therefore,counting the head number per unit area is critical for plant breeders to correlate with the genotypic variation in a specific breeding field.However,measuring such phenotypic traitsmanually is an extremely labor-intensive process and suffers from low efficiency and human errors.Moreover,the process is almost infeasible for large-scale breeding plantations or experiments.Machine learning-based approaches like deep convolutional neural network(CNN)based object detectors are promising tools for efficient object detection and counting.However,a significant limitation of such deep learningbased approaches is that they typically require a massive amount of hand-labeled images for training,which is still a tedious process.Here,we propose an active learning inspired weakly supervised deep learning framework for sorghum head detection and counting from UAV-based images.We demonstrate that it is possible to significantly reduce human labeling effort without compromising final model performance(R^(2)between human count and machine count is 0.88)by using a semitrained CNN model(i.e.,trained with limited labeled data)to perform synthetic annotation.In addition,we also visualize key features that the network learns.This improves trustworthiness by enabling users to better understand and trust the decisions that the trained deep learning model makes.