Skip to main content


Fig. 1 | Stem Cell Research & Therapy

Fig. 1

From: Single cell transcriptomics suggest that human adipocyte progenitor cells constitute a homogeneous cell population

Fig. 1

a First-level clustering of SVF cells from scWAT. Left: cell–cell (upper) and gene–gene (lower) distance matrices of cells and genes ordered according to cluster membership determined by first-level clustering. Pearson correlation used as distance metric. Right: heatmap showing normalized expression of genes (rows) over all cells (columns) in the dataset. Cells and genes ordered as shown on the left. Upper panel shows Patient ID membership of cells, while lower panel shows cluster membership. b t-SNE plot showing visualization of WAT cells in 2-dimensional space. Cells colored according to cluster membership introduced in (a). c Violin plots showing the most differentially expressed genes in each WAT cluster based on negative binominal regression analysis. A gene is defined as differentially expressed in a population if its posterior probability (PP) exceeds the PP of all other populations with at least 99% probability. Genes shown were selected from all significant differentially expressed genes according to distance between the median expression in the relevant population (colored violin) compared to second highest median expression in any other population (gray violin). d Expression of two top genes representing each cluster/t-SNE population (c) in flow cytometry-sorted WAT cell populations. Expression measurement performed by Affymetrix Clariom™D microarray, normalized values compared (n = 6). e t-SNE plots showing expression of selected WAT marker genes over the dataset. ATM adipose tissue macrophage, FACS fluorescence-activated cell sorting, t-SNE t-distributed stochastic neighbor embedding, hWAT human white adipose tissue

Back to article page