ezSingleCell-package: ezSingleCell: Interactive single-cell data analysis using the...

In MattMyint/ezSingleCell: Interactive single-cell data analysis via the Seurat pipeline

Description

This package was made to provide a GUI for carrying out single cell data analysis on Seurat, as opposed to running the entire pipeline from the command line. Analysis using the package starts after alignment and quantification of counts from raw reads, taking a sparse matrix of expression values.

Details

Data loading Upon loading count data, counts are log normalised, and filtered based on user input of minumum cut-offs, as well as expression thresholds. Users can also set cell population identities based on the formatting of cell names in their expression table.

Quality check plots Immediately after counts are loaded, plots visualising metadata such as nUMI and percentage of mitochondrial genes are generated to allow the user to determine if any cells are low quality and need to be filtered out from downstream analysis.

Variable gene identification Variable genes are identified with user-selected dispersion and mean cut-offs. Clicking "Find Variable genes" will return the number of variable genes identified for downstream analysis. This is sufficient to proceed with later steps (i.e. plotting the graph is not necessary). It should be noted that this step needs to be done so that dimension reduction and clustering can be done later on.

PCA PCA will be run on identified variable genes, and users can visualise 2D plots of selected PCs. This can provide better visualisation of any outliers. After PCA is done, users can search for clusters.

Diagnostic analysis of PCs with Jackstraw and Elbow plots Seurat's tSNE clustering outcomes are notably dependent on the PCs used, so users should take time to determine which PCs are significant for more accurate results.

tSNE dimension reduction tSNE will collapse the chosen PCs into lower dimensions and provide additional visualisation of clustering of cell populations.

Differentially Expressed Gene (DEG) analysis Users can identify differentially expressed genes across groups, either as a one-vs-all comparison, or a 1-on-1 comparison between 2 selected groups.

The generated figures can be saved to PDF and CSV files, and if the user wishes to export their current analysis, they can use the "Save Data" button to save the Seurat object as an .RObj file.

Author(s)

Matthew Myint

References

placeholder

MattMyint/ezSingleCell documentation built on May 8, 2019, 1:36 a.m.