Journal: NAR Genomics and Bioinformatics

Article Title: stPipe: a flexible and streamlined R/Bioconductor pipeline for preprocessing sequencing-based spatial transcriptomics data

doi: 10.1093/nargab/lqaf167

Figure Lengend Snippet: Interactive R-shiny web app created by the Run_Interactive function and visualization of QC metrics created by the Run_Visualization function for data quality assessment. ( a ) Visualization of Slide-seq mouse brain sample Puck_200115_08. The Run_Interactive function offers flexible options for selecting a ROI through four intuitive buttons. The ‘Add Selection’ button allows users to add spatial coordinates along with corresponding metadata, such as UMI count and spatial barcode sequences, each time an ROI is selected. The ‘Clear Last Selection’ button removes the most recently selected ROI from the current selection list. The ‘Reset All Selections’ button resets both the spatial heatmap and clustering plot, providing a clean slate for a new selection. Finally, the ‘Save All Selected ROI’ button saves the finalized selection as ‘selected_ROI’ object in the user’s R global environment, streamlining data management and export. In this example, the selection of cluster 7, highlighted in purple on the t-SNE plot, is found to mostly correspond to the choroid plexus region in the spatial UMI count plot. ( b ) Barplot showing spatial barcode demultiplexing information between 10× Visium probe-based (left) and polyA-based (right) protocols to assess sequencing accuracy. ( c ) Stacked bar plots showing the mapping rate, separated into reads that map to exons, introns, and those that are ambiguously mapped or map elsewhere in the genome (ordered by exon mapping rate) between 10× Visium probe-based (left) and polyA-based (right) protocols. ( d ) UMI duplication plot between a probe-based sample (left) with a higher UMI duplication number than a polyA-based one (right). A distribution skewed toward lower duplication values indicates higher library complexity and minimal redundancy, suggesting that the sequencing depth is well-matched to the diversity of the transcriptome. In contrast, a pronounced tail toward higher duplication values suggests substantial over-sequencing or PCR amplification biases, as many reads may originate from the same underlying transcript molecule. (e) UMI count distribution between sample 709 with two protocols, the first and last two are plotted as distribution of raw UMI count per spot and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} $\log _{10}$\end{document} UMI count per gene respectively.

Article Snippet: Spatial transcriptomics technology has developed rapidly in recent years, with various sequencing-based platforms such as 10× Visium, Slide-seq, and Stereo-seq becoming widely used by researchers.

Techniques: Selection, Sequencing, Amplification

Journal: PLOS Computational Biology

Article Title: Network models for bridging denoising and identifying spatial domains of spatially resolved transcriptomics

doi: 10.1371/journal.pcbi.1013867

Figure Lengend Snippet: (A) Visium SRT data of breast cancer annotated by pathologists consists of IDC (invasive ductal carcinoma), DCIS (ductal carcinoma in situ), LCIS (lobular carcinoma in situ), tumor edge, and healthy region. (B) Spatial domains identified by stACN (left), stLearn (middle), and stACN-Con (right). (C) Heatmap of Pearson correlation coefficient among domains (domain=13). (D) Visualization of topological structure of spatial domains for breast cancer data in cell affinity graph learned by stACN, where thickness of edges is proportional to edge weights. (E) Distribution density estimation of cells in IDC, DCIS/LCIS and Healthy domain in terms of the learned cell features, where x-axis denotes cell features, and Kolmogorov-Smirnov test is for significance (left), and Distributions of degree, betweennessand eigenvector of cells in IDC, DCIS/LCIS and Healthy domains identified by stACN (right), where p-value is calculated with Student’s t-test. (F) UMAP visualization of spatial domains identified by stACN (left) and stLearn (right), where dashed circle denotes mixed domains. (G) Hierarchical structure of domain 3 and 14 in SRT data (left), and topological structure of subnetwork induced by domain 3 and 14 in cell affinity graph (right). (H) Spatial distribution of expression of GSTM3 and TFF1 with regional annotation (left), and Violin plots of gene expression (right).

Article Snippet: For the sake of convenience, we use cells to represent measurement units in spatial transcriptomics, such as spots in the 10× Visium platform, which can be interchanged freely. stACN models and characterizes the structure of noisy SRT data by learning compatible cell features from clean cell networks with graph denoising.

Techniques: In Situ, Expressing, Gene Expression

Journal: PLOS Computational Biology

Article Title: Network models for bridging denoising and identifying spatial domains of spatially resolved transcriptomics

doi: 10.1371/journal.pcbi.1013867

Figure Lengend Snippet: (A) H&E images of mouse anterior and posterior brain datasets of 10 × Visium, which are horizontally aligned (left). The zoomed in region consists of cornu ammonis(CA) and dentate gyrus(DG) domain. The corresponding anatomical Allen Mouse Brain Atlas (right). (B) Spatial domains identified by stACN (left) and STAGATE (right), where CA and DG across different slices. (C) 3D coordinates of MERFISH data for mouse hypothalamic preoptic region with slice 4, 9, and 14 (left), and spatial domains identified by stACN for each slice (right). (D) Visualization of SRT data for mouse breast cancer, where slice S1 and S3 are from different batches (first two columns), visualization of slice S1 and S3 with and without removing batch effect (the third column), and spatial domains identified by stACN with and without removing batch effect (last two columns), respectively.

Article Snippet: For the sake of convenience, we use cells to represent measurement units in spatial transcriptomics, such as spots in the 10× Visium platform, which can be interchanged freely. stACN models and characterizes the structure of noisy SRT data by learning compatible cell features from clean cell networks with graph denoising.

Techniques:

Journal: mBio

Article Title: A spatial transcriptomic atlas of the host response to oropharyngeal candidiasis

doi: 10.1128/mbio.00849-25

Figure Lengend Snippet: Major cell type annotation. Spatially resolved transcriptome ( A ) and t-SNE plots ( B ) showing representation of five major cell types in the tongue tissue. ( C ) Cell Chat software revealed the number and strength of inferred cellular signaling interactions from the spatial transcriptomics data.

Article Snippet: To analyze the microenvironment during OPC, we employed the 10× Genomics Visium spatial transcriptomics technology on frozen tissue sections ( n = 4) from tongues of normal and C. albicans -infected Balb/c mice at 60 h of OPC (hereby referred to as day 2, for ease of representation).

Techniques: Software

Journal: Alzheimer's Research & Therapy

Article Title: Exploring cellular heterogeneity: single-cell and spatial transcriptomics of Alzheimer's disease brains and iPSC-derived microglia

doi: 10.1186/s13195-025-01944-y

Figure Lengend Snippet: Integration of Single-cell transcriptomics datasets. a Diagram depicting the single-cell transcriptomics dataset utilized. b Highlighted transcriptional states selected from each single-cell transcriptomics dataset, demarcated with dotted lines. c UMAP plot showing 14 distinct integrated clusters labeled 0–13, comprising a total of 222,822 cells. d Quantification of individual cell state contributions to the integrated transcriptional state. e – f Gene expression analysis within each transcriptional state, referencing studies by Yun Chen et al. and Sun Victor et al. . Cluster numbers and gene names are highlighted with the same color code to indicate enrichment. Note: Xenografted-mic term used for Xenografted-microglia

Article Snippet: Fig. 3 Microglial transcriptional shift in response to AD pathology. a Spatial transcriptomics (SRT) of the Middle Temporal Gyrus (MTG) in Alzheimer's disease (AD), with each section being 10 μm thick. b Visium spots highlighting the top 25% highest probability for Homeostatic, DAM, MHCII, Neuronal Surveillance and Inflammatory-I states. c Heatmap illustrating the fraction of predicted transcriptional states within each cortical layer. d Overview of spatial transcriptomics Aβ localization.

Techniques: Single-cell Transcriptomics, Labeling, Gene Expression

Journal: Alzheimer's Research & Therapy

Article Title: Exploring cellular heterogeneity: single-cell and spatial transcriptomics of Alzheimer's disease brains and iPSC-derived microglia

doi: 10.1186/s13195-025-01944-y

Figure Lengend Snippet: Microglial transcriptional shift in response to AD pathology. a Spatial transcriptomics (SRT) of the Middle Temporal Gyrus (MTG) in Alzheimer's disease (AD), with each section being 10 µm thick. b Visium spots highlighting the top 25% highest probability for Homeostatic, DAM, MHCII, Neuronal Surveillance and Inflammatory-I states. c Heatmap illustrating the fraction of predicted transcriptional states within each cortical layer. d Overview of spatial transcriptomics Aβ localization. Aβ-proximal spots refer to those directly overlapping Aβ plaques, while all others are considered Aβ-distal. e Upper: Quantification of transcriptional states around proximal and distal Aβ spots for Combined II-VI, External II-III, and Internal IV-VI cortical layers. f - j SRT sample from AD frontal cortex from van Olst et al. . f Spatially resolved clusters based on gene expression from van Olst et al. AD sample. g Cortical layers identified based on main layer markers reported in van Olst et al., shown in panel h . The grey matter layers were identified as External (Layers I-III) and Internal (Layers IV-VI). Meninges and white mater were not considered in the analysis. i The Homeostatic and DAM enriched spots identified across the grey matter. j Proportion of Homeostatic and DAM enriched spots in each Internal and External layers. Chi-square significance tests were used to calculate p-values (refer Fig. S9 for other transcriptional states). Note: Xenografted-mic term used for Xenografted-microglia

Article Snippet: Fig. 3 Microglial transcriptional shift in response to AD pathology. a Spatial transcriptomics (SRT) of the Middle Temporal Gyrus (MTG) in Alzheimer's disease (AD), with each section being 10 μm thick. b Visium spots highlighting the top 25% highest probability for Homeostatic, DAM, MHCII, Neuronal Surveillance and Inflammatory-I states. c Heatmap illustrating the fraction of predicted transcriptional states within each cortical layer. d Overview of spatial transcriptomics Aβ localization.

Techniques: Gene Expression

Journal: Alzheimer's Research & Therapy

Article Title: Exploring cellular heterogeneity: single-cell and spatial transcriptomics of Alzheimer's disease brains and iPSC-derived microglia

doi: 10.1186/s13195-025-01944-y

Figure Lengend Snippet: Spatial distribution of microglial activation across cortical layers in Alzheimer’s disease (AD) brain. a Immunofluorescence (IF) staining of P2RY12 and Aβ on adjacent Sects. (10 µm interval) from the Middle Temporal Gyrus of an AD donor, aligned to 10X Genomics Visium spatial transcriptomics spots (color-coded) across cortical layers II–VI (Chen et al., ANC, 2022) . High-magnification images show nuclei (DAPI, gray), homeostatic microglia (P2RY12, magenta), and Aβ plaques (blue) in external layers II–III (top) and internal layers IV–VI (bottom). b Quantification of IF-stained P2RY12⁺ cells and Aβ⁺ plaques across cortical layers II–III and IV–VI in AD samples. Bar plots display normalized counts for: Upper Left—P2RY12⁺ cells; Upper Right—Aβ⁺ plaques; Lower Left—P2RY12⁺/Aβ⁺ overlap; Lower Right—P2RY12⁺/Aβ⁻ plaques. Counts were normalized to the total number within layers II–VI. c IF co-staining of Aβ (red) and phosphorylated tau (pTAU, green) in frontal cortex sections with AD pathology (Section A). Nuclei stained with DAPI (blue). Adjacent section (Section B) stained for CD68 (red), a marker of activated microglia. d Quantification of CD68⁺ cells across cortical layers in AD frontal cortex. Graph shows distribution of CD68⁺ and CD68⁻ cells in external versus internal layers

Article Snippet: Fig. 3 Microglial transcriptional shift in response to AD pathology. a Spatial transcriptomics (SRT) of the Middle Temporal Gyrus (MTG) in Alzheimer's disease (AD), with each section being 10 μm thick. b Visium spots highlighting the top 25% highest probability for Homeostatic, DAM, MHCII, Neuronal Surveillance and Inflammatory-I states. c Heatmap illustrating the fraction of predicted transcriptional states within each cortical layer. d Overview of spatial transcriptomics Aβ localization.

Techniques: Activation Assay, Immunofluorescence, Staining, Marker