Concordance of MERFISH Spatial Transcriptomics with Bulk and Single-cell RNA Sequencing

Abstract

Spatial transcriptomics extends single cell RNA sequencing (scRNA-seq) technologies by providing spatial context for cell type identification and analysis. In particular, imaging-based spatial technologies such as Multiplexed Error-Robust Fluorescence In Situ Hybridization (MERFISH) can achieve single-cell resolution, allowing for the direct mapping of single cell identities to spatial positions. Nevertheless, because MERFISH produces an intrinsically different data type than scRNA-seq methods, a technical comparison between the two modalities is necessary to ascertain how best to integrate them. Here, we used the Vizgen MERSCOPE platform to perform MERFISH on mouse liver and kidney tissues and compared the resulting bulk and single-cell RNA statistics with those from the existing Tabula Muris Senis cell atlas. We found that MERFISH produced measurements that quantitatively reproduced the bulk RNA-seq and scRNA-seq results, with some minor differences in overall gene dropout rates and single-cell transcript count statistics. Finally, we explored the ability of MERFISH to identify cell types, and found that it could independently resolve distinct cell types and spatial structure in both liver and kidney. Computational integration with the Tabula Muris Senis atlas using scVI and scANVI did not noticeably enhance these results. We conclude that compared to scRNA-seq, MERFISH provides a quantitatively comparable method for measuring single-cell gene expression, and that efficient gene panel design allows for robust identification of cell types with intact spatial information without the need for computational integration with scRNA-seq reference atlases.

SiGra: Single-cell spatial elucidation through image-augmented graph transformer

“…Spatial transcriptomics data generated by different platforms, including the NanoString CosMx TM SMI lung cancer dataset (Lung-9-1) 4 , Vizgen MERSCOPE mouse liver dataset L1R1 released in January 2022 15 , and 10x Visium datasets from human dorsolateral prefrontal cortex (DLPFC) 50 , are preprocessed and represented in the uniform format (Fig. 1) for SiGra.…”

“…In addition to domain recognition, the enhancement of spatial gene expression data also presents a significant challenge. Though great progress has been made in spatial technologies, the major problems such as missing values, data sparsity, low coverage, and noises 2,15 encountered in spatial transcriptomics profiles are impeding the effective use and the elucidation of biology insights 16,17 . Meanwhile, the multi-channel spatial images in single-cell spatial data consist of high-resolution, high-content features detected in the tissue, such as cell types, functions, and morphologies of cellular compartments, as well as the spatial distributions of cells.…”

Predictive and robust gene selection for spatial transcriptomics

“…Importantly, because reference datasets consisting of FISH-based measurements for thousands of genes are unavailable, scRNA-seq datasets are used instead to guide the selection process. The use of a surrogate dataset presents new obstacles: the expression counts observed by scRNA-seq and FISH may differ significantly, with a relationship that is nonlinear and noisy [16, 17, 18]. Hence, a gene panel selected without considering the difference between the datasets is unlikely to perform as well as intended, as we demonstrate with several existing methods.…”

High-plex Multiomic Analysis in FFPE at Subcellular Level by Spatial Molecular Imaging

“…In addition, SMI data was compared to this scRNA-seq dataset for the fraction of cells that have non-zero counts for each of the 788 genes that are common in both datasets (Figure S7B). This fraction for each gene indicates the sensitivity ratio between the two technologies, as recently described in (39). The data points above the slope=1 line in a scatter plot would indicate that one technology has higher sensitivity than the other.…”