Quantitative morphology of RNAi-induced phenotypes in cultured cells

我们继续开发我们的高密度RNAi筛选技术，我们已经开始对来自小规模试点屏幕的图像数据进行表征。量化图像相似性的能力使我们能够证明，敲除已知具有强烈遗传或物理交互作用的基因会导致高度相似的表型。关键目标之一是能够直接比较来自独立筛查的表型。我们的初步数据表明，将独立的筛选结合到一个通用的分类器中，使其能够更准确地表示更广泛定义的表型类别(例如，细胞周期阶段的块)，同时保持因敲除已知的物理相互作用的蛋白质编码基因而产生的表型之间的非常接近的相似性。进一步改进这一筛选平台的工作仍在继续，以实现更大程度的一致性和更通用的分类器。 我们已经开始与Ko博士(LG-NIA)合作，利用转录因子操作来表征在小鼠胚胎干细胞中诱导的发育途径(1)。第一组试点实验将沿着以前已知的途径推动分化，作为一套训练控制。我们将使用WND-Charm来确定这些途径的分化时间进程，以及确定特定途径可以唯一识别的最早阶段。 我们已经开始工作，以确定新的表型(或发育途径)是否可以在没有先验定义的情况下通过WND-Charm来识别。初步证据表明，这是可能的非常不同的形态，如细胞凋亡，研究工作正在进行中，如果更微妙的表型变化也可以识别。在一项已发表的研究(2)中，我们能够证明WND-Charm可以分辨出由同一基因的不同突变等位基因引起的非常微妙的细胞表型。

We have continued to develop our high-density RNAi screening technology, and we have begun to characterize image data from small-scale pilot screens. The ability to quantify image similarity has allowed us to demonstrate that knock-down of genes known to have strong genetic or physical interactions leads to highly similar phenotypes. One of the key objectives was the ability to directly compare the phenotypes resulting from independent screens. Our preliminary data indicates that combining independent screens into a common classifier enables it to more accurately represent more broadly defined phenotypic classes (e.g. blocks in cell cycle stages), while maintaining the very close similarity between phenotypes resulting from knockdown of genes encoding proteins known to physically interact. Work continues on further refinement of this screening platform to achieve a greater degree of consistency and a more universal classifier. We have begun work on a collaboration with Dr. Minoru Ko (LG-NIA) to characterize developmental pathways induced in mouse embryonic stem cells using transcription factor manipulation (1). The first set of pilot experiments will be to drive differentiation along previously known pathways to serve as a set of training controls. We will use WND-CHARM to determine the time course of differentiation for these pathways as well as determine the earliest stages that a specific pathway can be uniquely identified. We have begun work to determine if new phenotypes (or developmental pathways) can be identified by WND-CHARM without them being defined a priori. Preliminary evidence suggests this is possible with very distinct morphologies such as apoptosis, and work is ongoing to investigate if more subtle phenotypic variations can also be discerned. In a published study (2), we were able to demonstrate that WND-CHARM can discern very subtle cellular phenotypes resulting from different mutant alleles of the same gene.