A collection of useful functions for scRNA-seq analysis, meant to be utilized with tiledbsoma experiment objects.
This R package was a personal project that I undertook during the winter of 2023-2024. The purpose of the project was to learn how to perform pseudo-bulking for scRNA-seq data, how to employ various differential expression analysis methods, and how to interact with tiledbsoma objects. This repository is no longer updated, and it is not representative of best practices that I currently use in my work.
The package requires the tiledbsoma package be installed. See this repository for details on this installation. Once this dependency is installed, one can install Catullus as follows:
remotes::install_github("ccnawrocki/Catullus")
There are three function types in Catullus: Do, View, and Get.
- A
Dofunction performs an operation or test. - A
Viewfunction creates a plot. - A
Getfunction simply retrieves something for future use.
Here is a list of all the functions included in Catullus:
DoDETesting
DoPseudobulkAggregation
ViewDETesting
ViewMeta
ViewExpression
GetDEData
GetObjectSchema
GetGeneData
GetCellGroupIDs
GetDimRedData
GetMetaData
GetExpressionData
The following vignette uses a subset of the data from Briggs et al. 2018.
First, open the tiledbsoma experiment object and check the schema.
library(Catullus)
experiment <- tiledbsoma::SOMAExperimentOpen("briggs_data_stage22_SOMA")
GetObjectSchema(experiment)
Examining the schema (not shown), we see that louvain clustering was performed on this data and that the results are saved in the metadata in the "louvain_33_clusters" column. Let's look at the UMAP, labeling the clusters.
Catullus::ViewMeta(exp_object = experiment, color_var = "louvain_33_clusters", labeling = T, legend = F)
Now, let's look again at the UMAP, but let's see which clusters express a known gene of interest, pax6.
Catullus::ViewExpression(exp_object = experiment, genes = "pax6", label_var = "louvain_33_clusters")
