3D computational reconstruction of tissues with hollow spherical morphologies using single-cell gene expression data.

Abstract

Single-cell gene expression analysis has contributed to a better understanding of the transcriptional heterogeneity in a variety of model systems, including those used in research in developmental, cancer and stem cell biology. Nowadays, technological advances facilitate the generation of large gene expression data sets in high-throughput format. Strategies are needed to pertinently visualize this information in a tissue structure-related context, so as to improve data analysis and aid the drawing of meaningful conclusions. Here we describe an approach that uses spatial properties of the tissue source to enable the reconstruction of hollow sphere-shaped tissues and organs from single-cell gene expression data in 3D space. To demonstrate our method, we used cells of the mouse otocyst and the renal vesicle as examples. This protocol presents a straightforward computational expression analysis workflow, and it is implemented on the MATLAB and R statistical computing and graphics software platforms. Hands-on time for typical experiments can be <1 h using a standard desktop PC or Mac.