RNAseq-Analyser
The scripts in this repository are used to analyse RNA-seq data using Seurat. individual.R allows you to analyse a single dataset, while compare.R lets you compare two datasets to uncover differential gene expression.
Both scripts expect dataset folders to include the following correctly formatted files: barcodes.tsv, genes.tsv, matrix.mtx. Take a look at the RNAseq-Parser repository if you need to format your input files.
The sample data referred to in this repository can be obtained from here. It must be processed using the RNAseq-Parser repository before use.
Usage
Download with:
git clone https://github.com/phenotypic/RNAseq-Analyser.git
To run the scripts you must have RStudio installed. Download through brew by running: brew install --cask rstudio
To install any outstanding libraries, run R from the command line, then run install.packages("package_name"), replacing package_name as necessary.
Results
After running individual.R, you will be presented with the following results:
| Analysis | Description | Figure |
|---|---|---|
| Quality control | Filter cells that have unique feature counts over 2,500 or less than 200, have >5% mitochondrial counts |  |
| Identify variable features | Features exhibiting high cell-to-cell variation in the dataset (i.e, they are highly expressed in some cells, and lowly expressed in others) |  |
| Linear dimension reduction | Perform principal component analysis (PCA) on the scaled data. Useful way to visualise cells and features that define the PCA |  |
| Dimensional heat map | Allows for easy exploration of the primary sources of heterogeneity in a dataset |  |
| Determine dimensionality | Comparing the distribution of p-values for each PC with a uniform distribution |  |
| Non-linear dimensional reduction | Visualise and explore dataset by grouping similar cells together in low-dimensional space |  |
compare.R will output lists of the most highly differentially expressed genes and generate some figures:
| Title | Figure |
|---|---|
| Integrated analysis of all cells: clustering |  |
| Visualising spatial information for the most differentially expressed genes |  |