Multiplexed DNA-FISH imaging dataset, drosophila embryos, nuclear cycles 11-14

Multiplexed DNA-FISH imaging dataset from Drosophila embryos at nuclear cycles 11-14. Examples on how to load and use this dataset can be found at this GitHub repository. Data processing details Barcodes were segmented using a neural network (stardist) specifically trained for the detection of 3D diffraction limited spots produced by our microscope. To extract the position of the barcode with sub-pixel accuracy, a subsequent 3D Gaussian fit of the regions segmented by stardist was performed with Big-FISH (https://github.com/fish-quant/big-fish). Barcode localizations with intensities lower than 1.5 times that of the background were filtered out. Nuclei were segmented from projected DAPI images using stardist with a neural network trained for detection of nuclei from Drosophila embryos under our imaging conditions. Barcodes were then attributed to single nuclei by using the XY coordinates of the barcodes and the DAPI masks of the nuclei. Finally, pairwise distance matrices were calculated for each single nucleus. Processed data in Figures This new version of the dataset contains the raw data for each of the figures in the manuscript: Associated publication Multiple parameters shape the 3D chromatin structure of single nuclei at the doc locus in Drosophila. Markus Götz, Olivier Messina, Sergio Espinola, Jean-Bernard Fiche, Marcelo Nollmann Nature Communications (2022).

果蝇胚胎11-14次核周期的多重DNA荧光原位杂交（DNA-FISH）成像数据集。该数据集的加载与使用示例可于本GitHub仓库中获取。
数据处理细节：采用专为检测本实验室显微镜所采集的三维衍射受限斑点而训练的神经网络（StarDist）对条形码信号进行分割；为提取条形码信号的亚像素级精准位置，借助Big-FISH工具（https://github.com/fish-quant/big-fish）对StarDist分割出的区域执行三维高斯拟合；信号强度低于背景1.5倍的条形码定位结果将被过滤移除。采用针对本实验室成像条件下果蝇胚胎细胞核检测而训练的StarDist神经网络，从投影后的DAPI图像中分割出细胞核；随后通过条形码的XY坐标与细胞核的DAPI掩膜，将条形码信号归属至单个细胞核；最终为每个单个细胞核计算得到两两距离矩阵。
论文附图对应处理后数据：本数据集的新版本包含手稿中所有附图对应的原始数据。
相关研究论文：《多参数调控果蝇doc位点单细胞核三维染色质结构》，作者Markus Götz、Olivier Messina、Sergio Espinola、Jean-Bernard Fiche、Marcelo Nollmann，发表于《自然·通讯》（2022年）。