| 1 | Increased long-distance and homo-trans interactions related to H3K27me3 in Arabidopsis hybrids显示文摘In plants,the genome structure of hybrids changes compared with their parents,but the effects of these changes in hybrids remain elusive.Comparing reciprocal crosses between Col×C24 and C24×Col in Arabidopsis using high-throughput chromosome conformation capture assay(Hi-C)analysis,we found that hybrid three-dimensional(3D)chromatin organization had more long-distance interactions relative to parents,and this was mainly located in promoter regions and enriched in genes with heterosis-related pathways.The interactions between euchromatin and heterochromatin were increased,and the compartment strength decreased in hybrids.In compartment domain(CD)boundaries,the distal interactions were more in hybrids than their parents.In the hybrids of CURLY LEAF(clf)mutants clfCol×clfC24and clfC24×clfCol,the heterosis phenotype was damaged,and the long-distance interactions in hybrids were fewer than in their parents with lower H3K27me3.ChIP-seq data revealed higher levels of H3K27me3 in the region adjacent to the CD boundary and the same interactional homo-trans sites in the wild-type(WT)hybrids,which may have led to more long-distance interactions.In addition,the differentially expressed genes(DEGs)located in the boundaries of CDs and loop regions changed obviously in WT,and the functional enrichment for DEGs was different between WT and clf in the longdistance interactions and loop regions.Our findings may therefore propose a new epigenetic explanation of heterosis in the Arabidopsis hybrids and provide new insights into crop breeding and yield increase. | Zhaoxu Gao Yanning Su Le Chang Guanzhong Jiao Yang Ou Mei Yang Chao Xu Pengtao Liu Zejia Wang Zewen Qi Wenwen Liu Linhua Sun Guangming He Xing Wang Deng Hang He | 2024 | Journal of Integrative Plant Biology2024，66，2： | 0 |
| 2 | Point cloud completion via structured feature maps using a feedback network显示文摘In this paper,we tackle the challenging problem of point cloud completion from the perspective of feature learning.Our key observation is that to recover the underlying structures as well as surface details,given partial input,a fundamental component is a good feature representation that can capture both global structure and local geometric details.We accordingly first propose FSNet,a feature structuring module that can adaptively aggregate point-wise features into a 2D structured feature map by learning multiple latent patterns from local regions.We then integrate FSNet into a coarse-to-fine pipeline for point cloud completion.Specifically,a 2D convolutional neural network is adopted to decode feature maps from FSNet into a coarse and complete point cloud.Next,a point cloud upsampling network is used to generate a dense point cloud from the partial input and the coarse intermediate output.To efficiently exploit local structures and enhance point distribution uniformity,we propose IFNet,a point upsampling module with a self-correction mechanism that can progressively refine details of the generated dense point cloud.We have conducted qualitative and quantitative experiments on ShapeNet,MVP,and KITTI datasets,which demonstrate that our method outperforms stateof-the-art point cloud completion approaches. | Zejia Su Haibin Huang Chongyang Ma Hui Huang Ruizhen Hu | 2023 | Computational Visual Media2023，9，1： | 0 |
您所在的机构暂未开通此部分的功能，请联系管理员