Spatially Variable Genes

BOOST-GP a novel Bayesian hierarchical model to analyze spatial transcriptomics data, with several unique characteristics. It models the zero-inflated and over-dispersed counts by deploying a zero-inflated negative binomial model that greatly increases model stability and robustness. Besides, the Bayesian inference framework allows us to borrow strength in parameter estimation in a de novo fashion. As a result, the proposed model shows competitive performances in accuracy and robustness over existing methods in both simulation studies and two real data applications.

The GPcounts package implements GP regression methods for modelling counts data using a negative binomial likelihood function. Computational efficiency is achieved through the use of variational Bayesian inference. The GP function models changes in the mean of the negative binomial likelihood through a logarithmic link function and the dispersion parameter is fitted by maximum likelihood. We validate the method on simulated time course data, showing better performance to identify changes in over-dispersed counts data than methods based on Gaussian or Poisson likelihoods.

The MoranI global spatial auto-correlation statistics evaluates whether features (i.e. genes) shows a pattern that is clustered, dispersed or random in the tissue are under consideration.

nnSVG identifies genes that vary in expression continuously across the entire tissue or within a priori defined spatial domains. It uses gene-specific estimates of length scale parameters within the Gaussian process models, and scales linearly with the number of spatial locations.

Single-cell gene expression data with positional information is critical to dissect mechanisms and architectures of multicellular organisms, but the potential is limited by the scalability of current data analysis strategies. Here, we present scGCO, a method based on fast optimization of hidden Markov Random Fields with graph cuts to identify spatially variable genes. Comparing to existing methods, scGCO delivers a superior performance with lower false positive rate and improved specificity, while demonstrates a more robust performance in the presence of noises. Critically, scGCO scales near linearly with inputs and demonstrates orders of magnitude better running time and memory requirement than existing methods, and could represent a valuable solution when spatial transcriptomics data grows into millions of data points and beyond..

This method assesses the degree of randomness exhibited by each transcript profile and rank them accordingly.

SOMDE uses self-organizing map to cluster neighboring cells into nodes, and then uses a Gaussian process to fit the node-level spatial gene expression to identify SVgenes. Experiments show that SOMDE is about 5 to 50 times faster than existing methods with comparable results. The adjustable resolution of SOMDE makes it the only method that can give results in about 5 min in large datasets of more than 20 000 sequencing sites.

To elucidate spatial gene expression variation, we present SpaGCN, a graph convolutional network approach that integrates gene expression, spatial location and histology in SRT data analysis. Through graph convolution, SpaGCN aggregates gene expression of each spot from its neighboring spots, which enables the identification of spatial domains with coherent expression and histology. The subsequent domain guided differential expression (DE) analysis then detects genes with enriched expression patterns in the identified domains. Analyzing seven SRT datasets using SpaGCN, we show it can detect genes with much more enriched spatial expression patterns than competing methods. Furthermore, genes detected by SpaGCN are transferrable and can be utilized to study spatial variation of gene expression in other datasets. SpaGCN is computationally fast, platform independent, making it a desirable tool for diverse SRT studies.

The tissue module (TM) was defined as an architectural area containing recurrent cellular communities executing specific biological functions at different tissue sites. However, the computational identification of TMs poses challenges owing to their various length scales, convoluted biological processes, not well-defined molecular features, and irregular spatial patterns. Here, we present a hypothesis-free graph Fourier transform model, SpaGFT, to characterize TMs. For the first time, SpaGFT transforms complex gene expression patterns into simple, but informative signals, leading to the accurate identification of spatially variable genes (SVGs) at a fast computational speed. Based on clustering the transformed signals of the SVGs, SpaGFT provides a novel computational framework for TM characterization. Three case studies were used to illustrate TM identities, the biological processes of convoluted TMs in the lymph node, and conserved TMs across multiple samples constituting the complex organ. The superior accuracy, scalability, and interpretability of SpaGFT indicate that it is a novel and powerful tool for the investigation of TMs to gain new insights into a variety of biological questions.

The depiction of in situ gene expression through spatial transcriptomics facilitates the inference of cell function mechanisms. To build spatial maps of transcriptomes, the first and crucial step is to identify spatially variable (SV) genes. However, current methods fall short in dealing with large-scale spatial transcriptomics data and may result in a high false positive rate due to the modeling of gene expression into parametric distributions. This paper introduces Spanve (https://github.com/zjupgx/Spanve), a non-parametric statistical approach based on modeling space dependence as a distance of two distributions for detecting SV genes. The high computing efficiency and accuracy of Spanve is demonstrated through comprehensive benchmarking. Additionally, Spanve can detect clustering-friendly SV genes and spatially variable co-expression, facilitating the identification of spatial tissue domains by an imputation.

SPARK builds upon a generalized linear spatial model (GLSM) with a variety of spatial kernels to accommodate count data. With a newly developed penalized quasi-likelihood (PQL) algorithm, SPARK is scalable to analyzing tens of thousands of genes across tens of thousands spatial locations.

Spatial transcriptomic studies are becoming increasingly common and large, posing important statistical and computational challenges for many analytic tasks. Here, we present SPARK-X, a non-parametric method for rapid and effective detection of spatially expressed genes in large spatial transcriptomic studies. SPARK-X not only produces effective type I error control and high power but also brings orders of magnitude computational savings. We apply SPARK-X to analyze three large datasets, one of which is only analyzable by SPARK-X. In these data, SPARK-X identifies many spatially expressed genes including those that are spatially expressed within the same cell type, revealing new biological insights.

SpatialDE decomposes expression variability into spatial and nonspatial components using two random effect terms: a spatial variance term that parametrizes gene expression covariance by pairwise distances of samples, and a noise term that models nonspatial variability.

Spatial transcriptomics is now a mature technology, allowing to assay gene expression changes in the histological context of complex tissues. A canonical analysis workflow starts with the identification of tissue zones that share similar expression profiles, followed by the detection of highly variable or spatially variable genes. Rapid increases in the scale and complexity of spatial transcriptomic datasets demand that these analysis steps are conducted in a consistent and integrated manner, a requirement that is not met by current methods. To address this, we here present SpatialDE2, which unifies the mapping of tissue zones and spatial variable gene detection as integrated software framework, while at the same time advancing current algorithms for both of these steps. Formulated in a Bayesian framework, the model accounts for the Poisson count noise, while simultaneously offering superior computational speed compared to previous methods. We validate SpatialDE2 using simulated data and illustrate its utility in the context of two real-world applications to the spatial transcriptomics profiles of the mouse brain and human endometrium.

A negative control method with random ranking of genes.

A positive control method with correct ranking of genes.

Kendall rank correlation coefficient measures the ordinal association between two measured quantities. The best score and upper bound is 1 (observations have an identical rank), while the lower bound is -1 (observations have a completely different rank).

This dataset is provided as part of the Technical Note: Post-Xenium In Situ Applications: Immunofluorescence, H&E, and Visium CytAssist Spatial Gene Expression (CG000709). Post-Xenium samples were compared to controls (samples not processed through the Xenium workflow) using 5 µm (FFPE) serial sections.

This dataset is provided as part of the Technical Note: Post-Xenium In Situ Applications: Immunofluorescence, H&E, and Visium CytAssist Spatial Gene Expression (CG000709). Post-Xenium samples were compared to controls (samples not processed through the Xenium workflow) using 10 µm fresh-frozen (FF) serial sections.

10X Genomics obtained fresh frozen human cerebellum tissue from BioIVT Asterand. The tissue was embedded and cryosectioned as described in Visium Spatial Protocols Tissue Preparation Guide (Demonstrated Protocol CG000240). Tissue sections of 10µm were placed on Visium Gene Expression slides and fixed and stained following Methanol Fixation, H&E Staining & Imaging for Visium Spatial Protocols (CG000160).

10x Genomics obtained FFPE human brain cancer tissue from Avaden Biosciences. The tissue was sectioned as described in the Visium CytAssist Spatial Gene Expression for FFPE - Tissue Preparation Guide (CG000518). Tissue section of 5 µm was placed on a standard glass slide, then stained following the Deparaffinization, H&E Staining, Imaging & Decrosslinking Demonstrated Protocol (CG000520). The glass slide with tissue section was processed via Visium CytAssist instrument to transfer analytes to a Visium CytAssist Spatial Gene Expression Slide v2, with 11 mm capture areas following the Visium CytAssist Spatial Gene Expression Reagent Kits User Guide (CG000495).

10X Genomics obtained fresh frozen human Invasive Lobular Carcinoma breast tissue from BioIVT Asterand. The tissue was embedded and cryosectioned as described in Visium Spatial Protocols Tissue Preparation Guide Demonstrated Protocol (CG000240). Tissue sections of 10µm were placed on Visium Gene Expression slides and fixed and stained following Methanol Fixation, H&E Staining & Imaging for Visium Spatial Protocols (CG000160).

10x Genomics obtained FFPE human breast tissue from BioIVT Asterand Human Tissue Specimens. The tissue was annotated with Ductal Carcinoma In Situ, Invasive Carcinoma. The tissue was sectioned as described in Visium Spatial Gene Expression for FFPE - Tissue Preparation Guide Demonstrated Protocol (CG000408). Tissue sections of 5 µm were placed on Visium Gene Expression slides, then stained following Deparaffinization, H&E Staining, Imaging & Decrosslinking Demonstrated Protocol (CG000409).

5 µm section from squamous cell carcinoma of human cervical cancer. FFPE tissue purchased from Discovery Life Sciences.

The tissue was sectioned as described in the Visium CytAssist Spatial Gene Expression for FFPE Tissue Preparation Guide (CG000518). Tissue section of 5 µm was placed on a standard glass slide, then stained following the Deparaffinization, H&E Staining, Imaging & Decrosslinking Demonstrated Protocol (CG000520). The glass slide with tissue section was processed via Visium CytAssist instrument to transfer analytes to a Visium CytAssist Spatial Gene Expression Slide v2, with 11 mm capture areas following the Visium CytAssist Spatial Gene Expression Reagent Kits User Guide (CG000495).

10x Genomics obtained fresh frozen human heart tissue from BioIVT Asterand. The tissue was embedded and cryosectioned as described in Visium Spatial Protocols - Tissue Preparation Guide Demonstrated Protocol (CG000240). Tissue sections of 10 µm thickness were placed on Visium Gene Expression Slides.

Frozen heart samples were embedded in OCT (Tissue-Tek) and cryosectioned (Thermo Cryostar). The 10-µm section was placed on the pre-chilled Optimization slides (Visium, 10X Genomics, PN-1000193) and the optimal lysis time was determined. The tissues were treated as recommended by 10X Genomics and the optimization procedure showed an optimal permeabilization time of 12 or 18 min of digestion and release of RNA from the tissue slide. Spatial gene expression slides (Visium, 10X Genomics, PN-1000187) were used for spatial transcriptomics following the Visium User Guides

Frozen heart samples were embedded in OCT (Tissue-Tek) and cryosectioned (Thermo Cryostar). The 10-µm section was placed on the pre-chilled Optimization slides (Visium, 10X Genomics, PN-1000193) and the optimal lysis time was determined. The tissues were treated as recommended by 10X Genomics and the optimization procedure showed an optimal permeabilization time of 12 or 18 min of digestion and release of RNA from the tissue slide. Spatial gene expression slides (Visium, 10X Genomics, PN-1000187) were used for spatial transcriptomics following the Visium User Guides

5 µm section from Human Intestinal Cancer. FFPE tissue purchased from BioIVT Asterand Human Tissue Specimens. Libraries were prepared following the Visium Spatial Gene Expression Reagent Kits for FFPE User Guide (CG000407 Rev A).

10x Genomics obtained FFPE human kidney tissue from Avaden Biosciences. The tissue was sectioned as described in the Visium CytAssist Spatial Gene Expression for FFPE – Tissue Preparation Guide (CG000518). Tissue section of 5 µm was placed on a standard glass slide, then stained following the Deparaffinization, H&E Staining, Imaging & Decrosslinking Demonstrated Protocol (CG000520). The glass slide with tissue section was processed via Visium CytAssist instrument to transfer analytes to a Visium CytAssist Spatial Gene Expression Slide v2, with 11 mm capture areas following the Visium CytAssist Spatial Gene Expression Reagent Kits User Guide (CG000495).

10x Genomics obtained FFPE human lung cancer tissue from Avaden Biosciences. The tissue was sectioned as described in the Visium CytAssist Spatial Gene Expression for FFPE Tissue Preparation Guide (CG000518). Tissue section of 5 µm was placed on a standard glass slide, then stained following the Deparaffinization, H&E Staining, Imaging & Decrosslinking Demonstrated Protocol (CG000520). The glass slide with tissue section was processed via Visium CytAssist instrument to transfer analytes to a Visium CytAssist Spatial Gene Expression Slide v2, with 11 mm capture areas following the Visium CytAssist Spatial Gene Expression Reagent Kits User Guide (CG000495).

10x Genomics obtained fresh frozen human lymph node from BioIVT Asterand. The tissue was embedded and cryosectioned as described in Visium Spatial Protocols - Tissue Preparation Guide Demonstrated Protocol (CG000240). Tissue sections of 10 µm thickness were placed on Visium Gene Expression Slides.

10x Genomics obtained FFPE human prostate tissue from Indivumed Human Tissue Specimens. The tissue was sectioned as described in Visium Spatial Gene Expression for FFPE – Tissue Preparation Guide Demonstrated Protocol (CG000408). Tissue sections of 5 µm were placed on Visium Gene Expression slides, then stained following Deparaffinization, H&E Staining, Imaging & Decrosslinking Demonstrated Protocol (CG000409).

10x Genomics obtained FFPE human prostate tissue from Indivumed Human Tissue Specimens. Original diagnosis with adenocarcinoma. The tissue was sectioned as described in Visium Spatial Gene Expression for FFPE Tissue Preparation Guide Demonstrated Protocol (CG000408). Tissue sections of 10 µm were placed on Visium Gene Expression slides, then stained following Deparaffinization, Decrosslinking, Immunofluorescence Staining & Imaging Demonstrated Protocol (CG000410).

10x Genomics obtained FFPE Human Melanoma tissue blocks from Avaden Biosciences. The tissue was sectioned as described in Visium CytAssist Spatial Gene Expression for FFPE Tissue Preparation Guide Demonstrated Protocol (CG000518). Tissue sections of 5 µm was placed on a standard glass slide, deparaffinized followed by immunofluorescence (IF) staining. Sections were coverslipped with 85% glycerol, imaged, decoverslipped, followed by dehydration & decrosslinking Demonstrated Protocol (CG000519). The glass slide with tissue section was processed via Visium CytAssist instrument to transfer analytes to a Visium CytAssist Spatial Gene Expression slide. The probe extension and library construction steps follow the standard Visium for FFPE workflow outside of the instrument.

FFPE Mouse Brain tissue blocks sectioned as described in Visium CytAssist Spatial Gene Expression for FFPE - Tissue Preparation Guide Demonstrated Protocol. The H&E stained glass slide with tissue section was processed via Visium CytAssist instrument to transfer analytes to a Visium CytAssist Spatial Gene Expression slide. The probe extension and library construction steps follow the standard Visium for FFPE workflow outside of the instrument. The H&E image was acquired using Olympus VS200 Slide Scanning Microscope. Sequencing depth was 53,497 reads per spot. Sequencing configuration: 28bp read 1 (16bp Visium spatial barcode, 12bp UMI), 90bp read 2 (transcript), 10bp i7 sample barcode and 10bp i5 sample barcode. Key metrics include: 2,310 spots detected under tissue; 6,736 median genes per spot; 24,862 median UMI counts per spot.

The tissue was sectioned as described in Visium CytAssist Spatial Gene Expression for FFPE Tissue Preparation Guide Demonstrated Protocol CG000518. Tissue sections of 5 µm was placed on a standard glass slide, and H&E-stained following deparaffinization. Sections were coverslipped with 85% glycerol, imaged, decoverslipped, followed by dehydration & decrosslinking (Demonstrated Protocol CG000520). The glass slide with the tissue section was processed with the Visium CytAssist instrument to transfer analytes to a Visium CytAssist Spatial Gene Expression slide (11 mm Capture Area). The probe extension and library construction steps follow the standard Visium for FFPE workflow outside of the instrument.

10x Genomics obtained fresh frozen mouse kidney tissue from BioIVT Asterand. The tissue was embedded and cryosectioned as described in Visium Spatial Protocols - Tissue Preparation Guide Demonstrated Protocol (CG000240). Tissue sections of 10 µm thickness from a slice of the coronal plane were placed on Visium Gene Expression slides, then stained following the Methanol Fixation, H&E Staining & Imaging Demonstrated Protocol (CG000160).

The tissue was embedded and cryosectioned as described in Visium Spatial Protocols Tissue Preparation Guide (Demonstrated Protocol CG000240). Tissue sections of 10µm were placed on Visium Gene Expression slides, then fixed and stained following Methanol Fixation, H&E Staining & Imaging for Visium Spatial Protocols (CG000160).

Gene expression library of an E10 whole mouse embryo tissue (brain in early-stage organogenesis) profiled using DBiT-seq.

Gene expression library of an E10 whole mouse embryo tissue (eye in early-stage organogenesis) profiled using DBiT-seq.

Gene expression library of an E11 whole mouse embryo tissue (lower body in early-stage organogenesis) profiled using DBiT-seq.

Gene expression library of an E10 whole mouse embryo tissue profiled using DBiT-seq.

Gene expression library of an E11 whole mouse embryo tissue profiled using DBiT-seq.

Gene expression library of an E11 whole mouse embryo tissue profiled using DBiT-seq.

Spatially resolved profiling of human cerebral cortex (middle temopral gyrus) replicate 1 using multiplexed error-robust fluorescence in situ hybridization (MERFISH) (250 gene panel).

Spatially resolved profiling of human cerebral cortex (middle temopral gyrus) replicate 1 using multiplexed error-robust fluorescence in situ hybridization (MERFISH) (4000 gene panel).

Spatially resolved profiling of human cerebral cortex (middle temopral gyrus) replicate 2 using multiplexed error-robust fluorescence in situ hybridization (MERFISH) (4000 gene panel).

Spatially resolved profiling of human cerebral cortex (middle temopral gyrus) replicate 3 using multiplexed error-robust fluorescence in situ hybridization (MERFISH) (4000 gene panel).

Spatially resolved profiling of mouse cerebral cortex (visual cortex (VIS), auditory cortex (AUD) and temporal association area (TEa) unexpanded sections) using multiplexed error-robust fluorescence in situ hybridization (MERFISH).

Sagittal sections from mouse embryo at the 8-12 ss was profiled by seqFISH.

Gene expression library of mouse cerebellum profiled using Slide-seq V2.

Gene expression library of Mouse cortex profiled using Slide-seq V2.

Gene expression library of mouse hippocampus puck profiled using Slide-seq V2.

Gene expression library of mouse olfactory bulk puck profiled using Slide-seq V2.

Gene expression library of mouse somatosensory cortex puck profiled using Slide-seq V2.

A 100 mm2 region of the human prefrontal cortex from a neurotypical donor aged 78 years was profiled by Slide-tags.

A metastatic melanoma sample was profiled by Slide-tags.

A human tonsil was profiled by Slide-tags.

Mouse embryo tonsil was profiled by Slide-tags.

3D architecture of cell types in visual cortex volumes.

3D architecture of cell types in visual cortex volumes.

Drosophila has long been a successful model organism in multiple biomedical fields. Spatial gene expression patterns are critical for the understanding of complex pathways and interactions, whereas temporal gene expression changes are vital for studying highly dynamic physiological activities. Systematic studies in Drosophila are still impeded by the lack of spatiotemporal transcriptomic information. Here, utilizing spatial enhanced resolution omics-sequencing (Stereo-seq), we dissected the spatiotemporal transcriptomic changes of developing Drosophila with high resolution and sensitivity. (Data from an embryo collected 14-16 h after egg laying)

Drosophila has long been a successful model organism in multiple biomedical fields. Spatial gene expression patterns are critical for the understanding of complex pathways and interactions, whereas temporal gene expression changes are vital for studying highly dynamic physiological activities. Systematic studies in Drosophila are still impeded by the lack of spatiotemporal transcriptomic information. Here, utilizing spatial enhanced resolution omics-sequencing (Stereo-seq), we dissected the spatiotemporal transcriptomic changes of developing Drosophila with high resolution and sensitivity. (Data from an embryo collected 14-16 h after egg laying)

Drosophila has long been a successful model organism in multiple biomedical fields. Spatial gene expression patterns are critical for the understanding of complex pathways and interactions, whereas temporal gene expression changes are vital for studying highly dynamic physiological activities. Systematic studies in Drosophila are still impeded by the lack of spatiotemporal transcriptomic information. Here, utilizing spatial enhanced resolution omics-sequencing (Stereo-seq), we dissected the spatiotemporal transcriptomic changes of developing Drosophila with high resolution and sensitivity. (Data from an embryo collected 14-16 h after egg laying)

Drosophila has long been a successful model organism in multiple biomedical fields. Spatial gene expression patterns are critical for the understanding of complex pathways and interactions, whereas temporal gene expression changes are vital for studying highly dynamic physiological activities. Systematic studies in Drosophila are still impeded by the lack of spatiotemporal transcriptomic information. Here, utilizing spatial enhanced resolution omics-sequencing (Stereo-seq), we dissected the spatiotemporal transcriptomic changes of developing Drosophila with high resolution and sensitivity. (Data from an embryo collected 14-16 h after egg laying)

Drosophila has long been a successful model organism in multiple biomedical fields. Spatial gene expression patterns are critical for the understanding of complex pathways and interactions, whereas temporal gene expression changes are vital for studying highly dynamic physiological activities. Systematic studies in Drosophila are still impeded by the lack of spatiotemporal transcriptomic information. Here, utilizing spatial enhanced resolution omics-sequencing (Stereo-seq), we dissected the spatiotemporal transcriptomic changes of developing Drosophila with high resolution and sensitivity. (Data from an embryo collected 14-16 h after egg laying)