Background Fatty liver disease causes huge economic losses in the poultry industry due to its high occurrence and lethality rate.Three-dimensional(3D)chromatin architecture takes part in disease processing by regulati...Background Fatty liver disease causes huge economic losses in the poultry industry due to its high occurrence and lethality rate.Three-dimensional(3D)chromatin architecture takes part in disease processing by regulating tran-scriptional reprogramming.The study is carried out to investigate the alterations of hepatic 3D genome and H3K27ac profiling in early fatty liver(FLS)and reveal their effect on hepatic transcriptional reprogramming in laying hens.Results Results show that FLS model is constructed with obvious phenotypes including hepatic visible lipid deposi-tion as well as higher total triglyceride and cholesterol in serum.A/B compartment switching,topologically associat-ing domain(TAD)and chromatin loop changes are identified by high-throughput/resolution chromosome conforma-tion capture(HiC)technology.Targeted genes of these alternations in hepatic 3D genome organization significantly enrich pathways related to lipid metabolism and hepatic damage.H3K27ac differential peaks and differential expres-sion genes(DEGs)identified through RNA-seq analysis are also enriched in these pathways.Notably,certain DEGs are found to correspond with changes in 3D chromatin structure and H3K27ac binding in their promoters.DNA motif analysis reveals that candidate transcription factors are implicated in regulating transcriptional reprogram-ming.Furthermore,disturbed folate metabolism is observed,as evidenced by lower folate levels and altered enzyme expression.Conclusion Our findings establish a link between transcriptional reprogramming changes and 3D chromatin struc-ture variations during early FLS formation,which provides candidate transcription factors and folate as targets for FLS prevention or treatment.展开更多
为了满足视觉机器人能够精准抓取平面零件的需求,提出一种加入深度学习算法的零件识别与定位方法。首先,利用YOLOv4-tiny目标检测算法对目标物体进行识别,并提取出感兴趣区域(Region of Interest,ROI)送入PSPnet语义分割网络中进一步提...为了满足视觉机器人能够精准抓取平面零件的需求,提出一种加入深度学习算法的零件识别与定位方法。首先,利用YOLOv4-tiny目标检测算法对目标物体进行识别,并提取出感兴趣区域(Region of Interest,ROI)送入PSPnet语义分割网络中进一步提取ROI。然后,将ROI区域进行亚像素级的模板匹配,并计算目标物体的深度信息。在目标物体中心坐标求解中,以ROI区域的最大内接圆的圆心作为目标物体的中心。最后,利用D-H法对机器人进行运动学解算,并进行抓取试验。实验结果表明:该方法的深度误差率大约为0.72%,视觉机器人抓取零件成功率达到91%,具有较高的定位精度和抓取成功率,可以满足实际工业分拣搬运需求。展开更多
基金funded by the National Science Foundation of China (32372910 and 32102567)the Program for Shaanxi Science&Technology (2022KJXX-13, 2023-YBNY-144, K3031223077 and 2022GD-TSLD-46–0302)
文摘Background Fatty liver disease causes huge economic losses in the poultry industry due to its high occurrence and lethality rate.Three-dimensional(3D)chromatin architecture takes part in disease processing by regulating tran-scriptional reprogramming.The study is carried out to investigate the alterations of hepatic 3D genome and H3K27ac profiling in early fatty liver(FLS)and reveal their effect on hepatic transcriptional reprogramming in laying hens.Results Results show that FLS model is constructed with obvious phenotypes including hepatic visible lipid deposi-tion as well as higher total triglyceride and cholesterol in serum.A/B compartment switching,topologically associat-ing domain(TAD)and chromatin loop changes are identified by high-throughput/resolution chromosome conforma-tion capture(HiC)technology.Targeted genes of these alternations in hepatic 3D genome organization significantly enrich pathways related to lipid metabolism and hepatic damage.H3K27ac differential peaks and differential expres-sion genes(DEGs)identified through RNA-seq analysis are also enriched in these pathways.Notably,certain DEGs are found to correspond with changes in 3D chromatin structure and H3K27ac binding in their promoters.DNA motif analysis reveals that candidate transcription factors are implicated in regulating transcriptional reprogram-ming.Furthermore,disturbed folate metabolism is observed,as evidenced by lower folate levels and altered enzyme expression.Conclusion Our findings establish a link between transcriptional reprogramming changes and 3D chromatin struc-ture variations during early FLS formation,which provides candidate transcription factors and folate as targets for FLS prevention or treatment.
文摘为了满足视觉机器人能够精准抓取平面零件的需求,提出一种加入深度学习算法的零件识别与定位方法。首先,利用YOLOv4-tiny目标检测算法对目标物体进行识别,并提取出感兴趣区域(Region of Interest,ROI)送入PSPnet语义分割网络中进一步提取ROI。然后,将ROI区域进行亚像素级的模板匹配,并计算目标物体的深度信息。在目标物体中心坐标求解中,以ROI区域的最大内接圆的圆心作为目标物体的中心。最后,利用D-H法对机器人进行运动学解算,并进行抓取试验。实验结果表明:该方法的深度误差率大约为0.72%,视觉机器人抓取零件成功率达到91%,具有较高的定位精度和抓取成功率,可以满足实际工业分拣搬运需求。
