Revolutionizing rice breeding: High-throughput imaging unlocks new insights in rice genomics and phenomics

Rice (Oryza sativa L.) serves as a staple food and a primary calorie source for more than half of the global population. Its critical role has sparked interest in exploring the genetic mechanisms underlying rice phenotypes. Recent strides in molecular biology, particularly advancements in next-generation technologies and high-density single nucleotide polymorphism (SNP) genotyping, have expedited functional genomics studies in rice. These advancements aim to correlate genotypes with phenotypes for precise rice breeding. However, traditional phenotyping methods, characterized by manual, time-consuming, and often destructive processes, have lagged behind the progress in genotyping techniques. In response, high-throughput, nondestructive plant phenotyping platforms have emerged, revolutionizing crop breeding research. These platforms enable dynamic and accurate analyses of plant growth and are increasingly incorporating artificial intelligence methods, such as machine learning, to enhance image processing for phenotype analysis. The current research focus is on customizing phenotype analysis pipelines with advanced image-processing algorithms, aiming to improve the accuracy and efficiency of phenotype analyses in rice and other crops.

In June 2023, Plant Phenomics published a research article entitled by "A Strategy for the Acquisition and Analysis of Image-Based Phenome in Rice during the Whole Growth Period ".

In this study, researchers developed a method for acquiring and analyzing image-based traits (i-traits) in rice throughout its growth stages using a high-throughput visible light imaging platform. The methodology involved utilizing SegNet, a machine learning model, to segment rice panicles, and the comparison of six models for predicting panicle and plant biomass. The quadratic model demonstrated superior accuracy in biomass prediction, supported by high R² values and successful cross-validation. The study delved into growth and senescence trends in rice, employing six growth models to evaluate plant and panicle development, with the quadratic model emerging as the most effective in detailing dynamic changes in growth and maturation. Furthermore, the research extended to yield prediction using i-traits, revealing that a select group of these traits could explain up to 84.8 % of yield variance, highlighting the potential of i-traits in improving yield predictions. The study also investigated variations in phenotypic traits across different latitudes and rice genotypes, uncovering significant correlations between traits and environmental adaptability, illustrating the impact of geographical factors on crop development. A critical component of the research was the integration of genome-wide association studies (GWAS) with i-traits, leading to the identification of numerous putative quantitative trait loci (QTLs). These QTLs, linked to traits like plant senescence and yield, pave the way for advanced genetic research in rice. The study also utilized principal component analysis (PCA) to further investigate genetic factors related to organ and temporal traits in rice, revealing a significant QTL on chromosome 3.

In conclusion, this study represents a significant advancement in rice phenomics, offering new insights into plant growth, environmental adaptation, and genetic factors influencing rice traits. The integration of high-throughput phenotyping and GWAS presents a novel approach in rice breeding research, potentially leading to the development of more resilient and productive rice varieties.

###

References

Authors

Zhixin Tang^1,2†, Zhuo Chen^1†, Yuan Gao¹, Ruxian Xue¹, Zedong Geng², Qingyun Bu³, Yanyan Wang¹, Xiaoqian Chen¹, Yuqiang Jiang^1, Fan Chen¹, Wanneng Yang^2*, and Weijuan Hu^1*

Affiliations

¹Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, Beijing 100101, China.

²National Key Laboratory of Crop Genetic Improvement, National Center of Plant Gene Research (Wuhan), Hubei Hongshan Laboratory, Huazhong Agricultural University, Wuhan 430070, China.  

³Northeast Institute of Geography and Agroecology, Key Laboratory of Soybean Molecular Design Breeding, Chinese Academy of Sciences, Harbin 150081, China.

†These authors contributed equally to this work.

About Weijuan Hu & Wanneng Yang

Weijuan Hu: She is an associate professor at the Institute of Genetics and Developmental Biology, Chinese Academy of Sciences.

Wanneng Yang: He is a professor in the National Key Laboratory of Crop Genetic Improvement at Huazhong Agricultural University. His major interest in recent years is crop phenomics and agricultural photonics. The researchareas include high-throughput rice phenotyping techniques, optical imaging, and computers inagriculture.