Creating a new pseudotime ordering of cells

Nygen Analytics utilises the pseudotime function from the Scarf package (Dhapola et al., 2022), which uses a memory efficient implementation of PBA algorithm (Weinreb et al. 2018, PNAS) to estimate a pseudotime ordering of cells. The pseudotime analysis is run in a supervised manner where the user provides a source and a sink.

Parameters

Step 1  Choosing a selection of cells

This will determine the group of cells which will be ordered for the pseudotime analysis. You can use the lasso tool to select clusters of cells, use saved selections or use all cells in the current analysis. Ideally, the selection should include the groups of cells that can be identified as a source, a sink and the transitioning cells between the source and the sink.

Step 2  Number of modules

The number of modules will determine how many modules or groups the cells will be divided into. This is used for the marker gene search, where potential marker genes for each module are identified.

Step 3  Select source and sink clusters

Source clusters are used to identify the group of cells at the beginning of a pseudotime ordering. These are usually stem cells, progenitors, precursor cells which can be considered cells found at early cell development. Sink clusters are the groups of cells that are considered to be towards the end of a cell development process, or differentiated cells which may be found at the end of the pseudotime.

‍

Pseudotime analysis output

Modules and marker genes

Cells will be grouped into the number of modules entered for the parameter when starting an analysis. The list of potential marker genes for each module can also be copied. Module 1 will show the group of genes with high expression in cells from early pseudotime and the genes in the final module with high expression from genes in late pseudotime.

Heatmap

The heatmap shows the gene expression of cells progressing through pseudotime. Hover the mouse pointer over the cells on the heatmap to reveal the name of the gene.

Visualisation on scatter plot

The expression pattern from the pseudotime analysis can be visualised on the main scatter plot with the UMAP. to check if it aligns with the clustering on the UMAP.

‍

‍