高内涵成像与数据分析

High content imaging and data analysis

在药物筛选领域和细胞生物学研究中,高内涵成像系统的应用越来越广泛。本文主要根据我们实际科研经验对高内涵成像系统使用中影响实验质量的三个方面进行归纳和建议。在样品准备阶段,需要考虑的因素有微孔板、细胞和荧光标记等。针对这些影响因素,文中介绍了一些有益于提高实验效率和图像质量的经验。在高内涵成像阶段,根据实验目的选择合适的成像模式和物镜,同时权衡考虑成像时间和成像视野数,以期在合理的时间内获取高质量的高内涵图像数据。最后是高内涵数据分析阶段,流程一般包括图像校正、图像分割、生物学参数的提取、生物学参数的聚类或分类分析以及最终的结果输出。目前常用单一生物学参数的Z'值对高内涵成像实验进行评价,部分高内涵成像实验探索利用多参数进行数据分析和评价。结合机器自学习功能,开发更复杂的多参数的数据分析方法,是未来高内涵数据分析的发展方向。

The application of high content imaging system has become more and more popular in drug discovery and cell biology research.This review mainly focuses on three aspects which will critically affect experiment quality.In the sample preparation period,microplate,cell and fluorescence label must be taken into account carefully.Several useful experiences in choosing these parameters are presented in this report.When capturing images,according to the purpose of the experiment,proper optical mode,objective lens,number of field of view and exposure time must be carefully set up to get high-quality images.Data analysis is very important for all of the high content imaging experiments.Generally,data analysis consists of image pre-process,image segmentation,biological parameters extracts,data cluster or classification and final results output.Z＇ factor is mostly used to evaluate the robustness and stability in a high content screening assay.Some assays even explored multi-parametric methods for data analysis and evaluation.In the future,taking advantage of machine learning algorithms,more complicated multi-dimensional data analysis methods may be developed.