Protocol for using MULTILAYER to reveal molecular tissue substructures from digitized spatial transcriptomes

Abstract

Summary Spatially resolved transcriptomics (SrT) allow researchers to explore organ/tissue architecture from the angle of the gene programs involved in their molecular complexity. Here, we describe the use of MULTILAYER to reveal molecular tissue substructures from the analysis of localized transcriptomes (defined as gexels). MULTILAYER can process low- and high-resolution SrT data but also perform comparative analyses within multiple SrT readouts. For complete details on the use and execution of this protocol, please refer to Moehlin et al., 2021 .

Three-dimensional molecular cartography of human cerebral organoids revealed by double-barcoded spatial transcriptomics

“…The outcome of this primary analysis is processed by our previously described tool, MULTILAYER, which is able to normalize the spatial read-count levels, identify differentially expressed genes in a local context, detect gene co-expression patterns, and perform molecular tissue substructure partitioning. 6 , 7 …”

“…2021 Sep 14; 2(4):100823. 7 https://doi.org/10.1016/j.xpro.2021.100823 https://github.com/SysFate/MULTILAYER ; Zenodo; https://doi.org/10.5281/zenodo.8180632 Scanning Software Microvisioneer manualWSI Microvisioneer Other DNA pico litter spotter Scienion Sciflexarrayer S3 …”

Spatial RNA sequencing methods show high resolution of single cell in cancer metastasis and the formation of tumor microenvironment

“…We have summarized some research methods used to improve ST imaging during the past 2 years, basically involving deconvolution software designed to evaluate the localization of transcriptome expression in ST data through calculations, including SpaGCN [65], MULTILAYER [66], STARCH [67], SPARK-X [68], DeepSpaCE [69], spatialGE [70], MISTy [71], and SpotClean [72]. These methods are mainly aimed at rare cell types existing in complex and multi-level tissue regions that may not be detected by ST, which is equivalent to deep in situ sequencing of some key areas in the whole tissue section.…”

A 3-dimensional molecular cartography of human cerebral organoids revealed by double-barcoded spatial transcriptomics

“…In addition to the molecular biology strategy for SrT assays, we have generated a bioinformatics pipeline targeting the retrieval of the Gibson sequence flanked by two molecular barcodes prior to the alignment to the reference genome, mapping of genomic features, and the reconstruction of a spatial coordinates matrix associated to the retrieved read-counts per transcripts ( Figure 1F ). The outcome of this primary analysis is processed by our previously described tool MULTILAYER, able to normalize the spatial read-count levels, identify differentially expressed genes in a local context, detect gene co-expression patterns and perform molecular tissue substructure partitioning 6,7 .…”

“…The "tissue-focused" matrix presenting read counts associated to spatial coordinates in columns, known transcripts in rows was processed with our previously described tool MULTILAYER 6,7 . For human brain organoids processed in Figure 1, tissue partitioning has been performed with ≥ 4 contiguous gexels per pattern; Tanimoto Similarity threshold: 25%; while for the 3-dimensional tissue reconstruction in Figure 2, tissue partitioning has been performed with ≥ 3 contiguous gexels per pattern; Tanimoto Similarity threshold: 10%.…”