SCenic Viewer plot descriptions: what they are and what the results mean
SCenic Viewer lets you visualize your dataset with five interactive plot types. This page explains what each one shows, when to use it, and how to customize it. Every visual can be downloaded as a high-resolution JPEG, PNG, SVG, or PDF image.
UMAP plots
UMAP, or Uniform Manifold Approximation and Projection, is a dimensionality-reduction technique used in machine learning and data visualization. It is particularly popular for visualizing high-dimensional data in a lower-dimensional space. UMAP aims to preserve the local structure of the data: data points that are close to each other in the high-dimensional space should also be close in the lower-dimensional representation.
In the Seurat package, UMAP is generated as default, and the UMAP embeddings are used for graph generation. UMAPs are preferred over t-distributed stochastic neighbor embedding (tSNE) because they better preserve large-scale structures and are more consistent across different runs.
Generate a umap plot if you want to...
- Visualize high-dimensional data in 2D/3D for dimensionality reduction and clustering.
- Compare cell populations across experimental conditions, revealing how treatments or disease states impact cellular composition and function.
- Identify cell types or states: points (cells) that form distinct groups often represent different cell types or states, and the separation between clusters indicates significantly different gene expression profiles.
What the plot will tell you
- The overall layout of the UMAP plot can give insights into the relationships between major cell populations.
- Groups of similar cells will cluster together, revealing potential cell types.
How to customize the plot
- Labels: Customize the plot by adjusting plotting features, including labeling, axis titles, and font size.
- Colors: Select a color palette to color the plot. The plot defaults to a polychrome color palette.
- Groupings: Choose a grouping method to color the cells by Seurat clusters or predefined cell annotations.
- Treatment Split: Select the treatment groupings to facet the UMAP plot and enable comparison of cellular responses across different experimental conditions.
Violin plots
A violin plot combines aspects of a box plot and a kernel density plot. It depicts the distribution and probability density of RNA expression of cells across different categories or groups. Instead of traditional box-and-whisker representations, violin plots use a series of symmetrical kernel density plots (mirrored on either side of the axis) to show the distribution of data within each category. Wider sections indicate higher data density, narrower sections lower density.
Violin plots are particularly useful when you want to visualize the distribution of data and compare it across different groups or categories, offering insights into the shape, spread, and central tendency of the variable.
Generate a violin plot if you want to...
- Identify and visualize marker genes that are differentially expressed in specific cell types or states.
- Visualize the distribution of expression of a gene across different cell populations and treatment conditions.
- Compare the expression level and variability of a gene between groups, and assess the impact of a treatment on gene expression.
- Assess the heterogeneity within a cell population by examining the spread and shape of the violins. Wide or multimodal distributions suggest the presence of subpopulations or varying cell states.
What the plot will tell you
- The median, quartiles, and overall distribution of the data for each group.
- Wider violin shapes indicate higher variability.
- Differences in median position suggest changes in expression level.
How to customize the plot
- Labels: Adjust labeling, axis titles, and font size.
- Groupings: Choose a grouping method to color the cells by Seurat clusters or predefined cell annotations.
- Treatment Split: Select the treatment groupings to facet the plot and compare cellular responses across different experimental conditions.
- Cell Cluster: Limit the plot to the distribution of the gene expression level in one cell type of interest only.
Feature plots
A feature plot is a visualization used to display the expression of specific genes across individual cells in a single-cell RNA-seq experiment, drawn in a two-dimensional UMAP format. It helps in exploring and understanding the distribution of gene expression within a cell population.
You can select genes of interest, and the feature plot visualizes the distribution of their expression levels across the entire cell population. Feature plots can be generated for specific groups or clusters within the dataset, facilitating the identification of gene expression patterns associated with different conditions or cell types.
Generate a feature plot if you want to...
- Visualize the expression of a specific gene or set of genes in low-dimensional space (often alongside a UMAP plot).
- Observe unexpected cluster expression patterns so you can form new hypotheses about cellular functions, interactions, or regulatory mechanisms for further testing.
- Identify regions or clusters where the gene of interest shows high or low expression, indicating that the gene is specifically (in)active in certain cell types or states.
What the plot will tell you
- The relative expression level of a gene (usually color-coded) for each cell on the UMAP plot.
How to customize the plot
- Labels: Adjust labeling, axis titles, and font size.
- Quantile Cutoff: Adjust the upper and lower quantile cutoffs to filter gene expression levels displayed in the plot, focusing on specific expression ranges by excluding extreme values.
- Treatment Split: Select a treatment type to facet the UMAP plot, allowing comparison of cellular distributions and gene expression profiles across different experimental conditions.
Heatmap plots
In single-cell visualization, a heatmap plot is a graphical representation of gene expression data across multiple cells. It displays a matrix where rows typically represent genes, columns represent individual cells or cell types, and the color intensity within each cell indicates the level of expression for that specific gene.
Generate a heatmap plot if you want to...
- Visualize the expression levels of many genes across different samples or conditions.
- Compare gene expression profiles between healthy and diseased samples, highlighting differentially expressed genes that could be potential biomarkers or therapeutic targets.
- Explore relationships between genes and sample groups, and identify clusters of genes with similar expression patterns in a cell population.
- Identify genes that are differentially expressed between conditions.
- Identify technical artifacts or batch effects. Unusual patterns may indicate issues that need addressing in the data preprocessing steps.
What the plot will tell you
- A color gradient representing gene expression levels: red indicates high expression, blue indicates low expression.
- Similar color patterns across rows show co-expressed genes.
- Distinct color patterns between conditions (columns) show differentially expressed genes.
- Dendrograms alongside the heatmap indicate hierarchical clustering of genes (rows) and cell populations (columns), showing which genes or cells have similar expression patterns.
- Groups of genes or cells with similar expression profiles appear as blocks or stripes of similar colors.
How to customize the plot
- Clustering: Reorder rows and columns by similarity to surface expression blocks.
- Annotation Group: Select an annotation group to define the categories displayed along the X-axis of the heatmap.
- Cell Clusters: Limit the plot to visualize expression levels for one cell type of interest only.
- Treatment Split: Choose a treatment type to further divide the X-axis based on the applied treatment conditions, enabling a detailed comparison of gene expression levels within each cell group across treatments.
- Colors: Define a custom color palette by which to color the plot.
- Labels: Adjust labeling, axis titles, and font size.
Dot plots
In single-cell analysis, a dot plot is a visualization technique used to display gene expression data across different cell types or clusters. It's a compact and informative way to show both the proportion of cells expressing a gene (dot size) and the average expression level of the gene within those cells (dot color).
Generate a dot plot if you want to...
- Analyze the proportion of cells expressing a particular gene within each group (e.g., cell types, treatment conditions).
- Visualize expression patterns for multiple genes across different cell populations and states simultaneously — particularly useful for identifying co-expression patterns and understanding the relationships between genes.
- Identify groups with high or low percentages of cells expressing a gene of interest.
- Provide a quantitative view of gene expression, showing both the percentage of cells expressing each gene and the average expression level within those expressing cells.
What the plot will tell you
- The percentage of cells expressing the gene in a group or cell label, through dot size.
- The average expression level within a group, through dot color.
How to customize the plot
- Labels: Adjust labeling, axis titles, and font size.
- Colors: Define a custom color palette by which to color the plot.
Want to see what these plots look like on real data?
Browse the sample reports for real single-cell and differential expression output, or watch the video tutorials to generate these plots on your own data.