针对依赖人工主观判断水稻叶瘟抗性费时费力且准确率低的问题,本文提出了一种基于水稻冠层尺度RGB图像和掩膜区域卷积神经网络(mask regions with convolutional neural network,Mask-RCNN)深度学习框架的水稻叶瘟病斑识别检测方法,通...针对依赖人工主观判断水稻叶瘟抗性费时费力且准确率低的问题,本文提出了一种基于水稻冠层尺度RGB图像和掩膜区域卷积神经网络(mask regions with convolutional neural network,Mask-RCNN)深度学习框架的水稻叶瘟病斑识别检测方法,通过分析水稻RGB图像中不同类别病斑的数量信息,构建多种分类模型来评估病斑数量和抗性水平之间的关联性。首先采集包括粳稻品系、早籼品系和籼型恢复系等不同品系的水稻育种材料的苗期RGB图像,然后通过对输入图像进行预处理和标记,最终建立了用于识别水稻叶瘟病斑的Mask-RCNN模型,实现了叶瘟病斑的矩形框检测、掩膜分割和分类,其平均交并比(mean intersection over union,mIoU)为0.603。当采用0.5的交并比(intersection over union,IoU)阈值时,测试数据集的病斑检测平均准确率均值(mean average precision,mAP)为0.716。在基于病斑数量的抗性评估模型中,高斯过程支持向量机在测试数据集上取得了94.30%的最高抗性评估准确率。研究结果表明,基于水稻冠层RGB图像和Mask-RCNN模型可实现水稻叶瘟病的准确识别,检测的病斑数量特征和叶瘟抗性水平高度相关。本研究为水稻抗病性品种的高效选育提供了技术支撑。展开更多
Transcriptional regulation of cold-responsive genes plays crucial roles in plant cold tolerance,but the transcription factors(TFs)-centered regulatory networks remain largely unclear.In this study,we show that Monocul...Transcriptional regulation of cold-responsive genes plays crucial roles in plant cold tolerance,but the transcription factors(TFs)-centered regulatory networks remain largely unclear.In this study,we show that Monoculm1(MOC1),a critical TF controlling tiller number and plant height in rice,positively regulates rice cold tolerance at the seedling stage.We found that OsMPK4,a mitogen-activated protein kinase,phosphorylates and stabilizes MOC1 under cold stress.Further investigations revealed that MOC1 recruits the TFs OsbZIP79 and OsNAC5 to form a triple complex and subsequently enhances their stability by inhibiting proteasome-mediated degradation under cold stress.Notably,we found that the OsbZIP79-MOC1-OsNAC5 complex activates several cold-responsive genes,including Dehydration-responsive element-binding factor 1G(OsDREB1G),to confer rice cold tolerance.Haplotype analysis of the OsDREB1G promoter in>10,000 rice accessions identified the favorable haplotype and key variants that endow rice cold tolerance.Collectively,our work demonstrates a pivotal role of the OsMPK4-OsbZIP79-MOC1-OsNAC5-OsDREB1G module in regulating rice cold tolerance and provides genetic targets for improving cold tolerance through molecular breeding.展开更多
文摘针对依赖人工主观判断水稻叶瘟抗性费时费力且准确率低的问题,本文提出了一种基于水稻冠层尺度RGB图像和掩膜区域卷积神经网络(mask regions with convolutional neural network,Mask-RCNN)深度学习框架的水稻叶瘟病斑识别检测方法,通过分析水稻RGB图像中不同类别病斑的数量信息,构建多种分类模型来评估病斑数量和抗性水平之间的关联性。首先采集包括粳稻品系、早籼品系和籼型恢复系等不同品系的水稻育种材料的苗期RGB图像,然后通过对输入图像进行预处理和标记,最终建立了用于识别水稻叶瘟病斑的Mask-RCNN模型,实现了叶瘟病斑的矩形框检测、掩膜分割和分类,其平均交并比(mean intersection over union,mIoU)为0.603。当采用0.5的交并比(intersection over union,IoU)阈值时,测试数据集的病斑检测平均准确率均值(mean average precision,mAP)为0.716。在基于病斑数量的抗性评估模型中,高斯过程支持向量机在测试数据集上取得了94.30%的最高抗性评估准确率。研究结果表明,基于水稻冠层RGB图像和Mask-RCNN模型可实现水稻叶瘟病的准确识别,检测的病斑数量特征和叶瘟抗性水平高度相关。本研究为水稻抗病性品种的高效选育提供了技术支撑。
基金supported by the National Natural Science Foundation of China(32570362,32460077,and 32570749)the Hainan Province Science and Technology Special Fund(ZDYF2025XDNY093)+4 种基金the Hainan Provincial Natural Science Foundation of China(325RC799)the Nanhai New Star Technology Innovation Talent Platform Project of Hainan Province(NHXXRCXM202362)the Research Startup Funding from Hainan Institute of Zhejiang University(0201-6602-A12203)the Open Project Program of the State Key Laboratory of Rice Biology and Breeding(20240101)the Open Project Program of Jiaxing Academy of Agricultural Sciences(202405).
文摘Transcriptional regulation of cold-responsive genes plays crucial roles in plant cold tolerance,but the transcription factors(TFs)-centered regulatory networks remain largely unclear.In this study,we show that Monoculm1(MOC1),a critical TF controlling tiller number and plant height in rice,positively regulates rice cold tolerance at the seedling stage.We found that OsMPK4,a mitogen-activated protein kinase,phosphorylates and stabilizes MOC1 under cold stress.Further investigations revealed that MOC1 recruits the TFs OsbZIP79 and OsNAC5 to form a triple complex and subsequently enhances their stability by inhibiting proteasome-mediated degradation under cold stress.Notably,we found that the OsbZIP79-MOC1-OsNAC5 complex activates several cold-responsive genes,including Dehydration-responsive element-binding factor 1G(OsDREB1G),to confer rice cold tolerance.Haplotype analysis of the OsDREB1G promoter in>10,000 rice accessions identified the favorable haplotype and key variants that endow rice cold tolerance.Collectively,our work demonstrates a pivotal role of the OsMPK4-OsbZIP79-MOC1-OsNAC5-OsDREB1G module in regulating rice cold tolerance and provides genetic targets for improving cold tolerance through molecular breeding.
