starfish: scalable pipelines for image-based transcriptomics
Shannon Axelrod, Matthew Cai, Ambrose Carr, Jeremy Freeman, Deep Ganguli, Justin Kiggins, Brian Long, Tony Tung, Kevin A. Yamauchi
Abstract
The exploding field of single cell transcriptomics has begun to enable deep analysis of gene expression and cell types, but spatial context is lost in the preparation of tissue for these assays. Recent developments in biochemistry, microfluidics, and microscopy have come together to bring about an "alphabet soup" of technologies that enable sampling gene expression in situ, with varying levels of spatial resolution, sensitivity, and genetic depth. These technologies promise to permit biologists to ask new questions about the spatial relationships between cell type and interactions between gene expression and cell morphology. However, these assays generate very large microscopy datasets which are challenging to process using general microscopy analysis tools. Furthermore, many of these assays require specialized analysis to decode gene expression from multiplexed experimental designs.