Journal: bioRxiv

Article Title: Scalable longitudinal imaging and transcriptomics of cells in dynamic enclosures

doi: 10.64898/2026.05.05.723030

Figure Lengend Snippet: A: UMAP based on transcriptomic data from primary human preadipocytes differentiated for seven days on a fibronectin-coated flow cell. The colors correspond to different clusters based on transcriptomic analysis. B: Transcriptomic UMAP colored by the lipid accumulation score, defined as the ratio between the BODIPY stain and the nuclear stain in each CCE. The insets show examples of cells that are very close in gene expression space but differ in their lipid content. C: violin plots depicting the distribution of lipid accumulation scores (y axis) across the transcriptomic clusters (x axis). D: actual (x axis) vs predicted (y axis) lipid accumulation scores from the elastic net model. The plot is for the held-out test set (20% of the total data). E: Euler diagram showing the overlap between top-20 differentially expressed genes between transcriptomic clusters (blue) and model-selected predictors of lipid accumulation (pink). F: average Log2 fold-change between clusters (x axis) vs absolute model coefficient (y axis) for the genes selected by the model. The red color indicates genes that are among the top-20 differentially expressed genes between transcriptomic clusters. G: Gene expression UMAP colored by the top-3 positive predictors identified by the model, showing that the expression values of these genes are uniformly distributed across the UMAP based on global transcriptomic differences. H: UMAP based on transcriptomic data for BV2 mouse microglial cells. The colors correspond to different clusters based on transcriptomic analysis. I: transcriptomic UMAP colored by phagocytic activity as measured by pHrodo™ intensity after four hours. J: UMAP based on DINOv2 features, colored by phagocytic activity showing a greater degree of separation between high vs low phagocytic scores, compared to the transcriptomic UMAP in panel H. K: violin plots depicting the distribution of phagocytic scores (y axis) across the transcriptomic clusters (x axis). L: R 2 performance of elastic net models trained on expression-only features, DINOv2-only features or a combination of the two (x axis). The data refers to the held-out test set (20% of the total data). M: actual (x axis) vs predicted (y axis) phagocytic scores from the elastic net model using the combined expression and DINOv2 features. The plot is for the held-out test set (20% of the total data). N: Euler plot showing the overlap between top-20 differentially expressed genes between transcriptomic clusters (blue) and model-selected predictors of phagocytic activity (pink). O: average Log2 fold-change between clusters (x axis) vs absolute models coefficient (y axis) for the genes selected by the expression-only model. The red color indicates genes that are among the top-20 differentially expressed genes between transcriptomic clusters. P: ridge plots displaying the expression level (x axis) of Gpnmb and Clec4e across transcriptomic clusters (x axis). These two genes are among the top positive predictors for the gene expression-based model and have clear mechanistic evidence linking them to the phagocytosis process. However, their expression is very similar across all the transcriptomic clusters.

Article Snippet: Adipogenesis was induced using Adipocytes Differentiation Toolkit for Adipose Derived MSCs and Preadipocytes (ATCC, # PCS-500-050).

Techniques: Staining, Gene Expression, Expressing, Activity Assay

Journal: Neural Regeneration Research

Article Title: Single-cell RNA sequencing of the post–spinal cord injury dorsal root ganglia in cynomolgus monkeys: Elucidation of the cellular immune microenvironment of the central nervous system

doi: 10.4103/NRR.NRR-D-24-00974

Figure Lengend Snippet: Single-cell dataset with reduced dimension clustering and cell type identification in DRG cells from a cynomolgus monkey with SCI. (A) Schematic representation of the experimental workflow. (B) Cluster analysis of cell groups, and t-SNE distribution showing cluster analysis groupings. Each point represents a cell, and cells that are close in distance are considered to be of the same type. Different groups of cells are distinguished by different colors and numbers. (C) t-SNE plot of all cells collected from the DRG following SCI. Cells are colored and annotated by cell type. (D) Heatmap of distinct genes with high and low expression enriched in each cell type. (E) t-SNE distribution of different cell types in the DRG and spinal cord following SCI. DRG: Dorsal root ganglia; SCI: spinal cord injury; t-SNE: t-distributed stochastic neighbor embedding.

Article Snippet: For the 10X Genomics data, the Cell Ranger Single-Cell toolkit (v3.0.0; 10X Genomics, Majorbio Co., Ltd., Shanghai, China) was employed to align the reads and generate the gene–cell unique molecular identifier (UMI) matrix for each sample ( https://support.10xgenomics.com/single-cell-gene-expression/software/downloads/latest ).

Techniques: Single Cell, Expressing