BigSur is a tool for single cell RNA sequencing analysis which interfaces with Seurat objects to both
- Select statistically significant features to use in clustering.
- Identify statistically significant correlations between genes.
The principles surrounding these analyses are described in "Leveraging gene correlations in single cell transcriptomic data".
BigSurR can be installed directly from github using the devtools package.
devtools::install_github("landerlabcode/BigSurR")
Note, Seurat recently changed the structure of their assay objects. BigSur's feature selection process will no longer work on versions below 5.1.0 (Assay5 class required).
To use BigSur, first create a Seurat object containing the raw transcript counts for each gene. As a preprocessing step, ensure that there are no genes which have zero counts across all cells.
library(BigSur)
example.seurat <- CreateSeuratObject(data, min.cells=1)
Pass this object into the BigSur function with the desired parameters. The parameters are as follows:
- seurat.obj: Seurat object containing the raw transcript counts filtered for zero count genes.
- assay: Assay slot containing raw transcript counts (default "RNA").
- counts.slot: Slot within assay containing raw counts matrix (default "counts").
- c: Boolean, double. The coefficient of variation of the dataset. If set to false, one will be estimated using the dataset.
- variable.features: Boolean. If true, BigSur will identify select variable features based on the modified corrected Fano factor. (Default true)
- correlations:Boolean. If true, BigSur will identify statistically significant gene-gene correlations. (Default false)
- first.pass.cutoff: Integer. Removes roots before p-value calculations if the root is below Abs[Sqrt(2)*InverseErfc(2*10^-first.pass.cutoff)]. The higher the number, the more correlations are removed in initial screening.
- inverse.fano.moments: Boolean. If true, BigSur will calculate the moments for the inverse Fano factor pairs before performing Cornish Fisher expansion. Setting false will significantly speed up calculation at the cost of some accuracy.
- fano.alpha: Double. Desired false discovery cutoff for labeling of variable features. (Default 0.05).
- min.fano: Double. Minimum mcFano value considered for variable genes.
- cor.alpha: Double. Desired false discovery cutoff for labeling of statistically significant correlations.
- depthlist: Boolean, vector. If a vector is supplied, that vector of values will be used to scale counts to account for unequal sequencing depth across cells. If left as False, this scaling will be calculated during the BigSur run.
- return.ps: Boolean. If true, the Benjamini-Hochberg corrected p-values associated with each equivalent PCC will be returned in a list with the equivalent PCC sparse matrix. The first object in this list will be the equivalent PCCs, the second will be the p-value matrix.
- log.file: Boolean. If true, a log file will be created.
- log.file.dir: String. Path of desired location for log file. *Note: Default string is set to work on Unix based file structures (i.e., manually set this on Windows).
For example, to calculate both the highly variable features and the significant correlations in your dataset with default false discovery cutoffs you could use:
example.output <- BigSur(example.seurat, variable.feature=T, correlations=T)
If both variable features and correlations are identified, a list containing the updated Seurat object and the an adjacency matrix of the statistically significant correlations is returned. If only one process is selected, their respective output is returned alone.
Using the entire dataset is very memory and computationally expensive. Often times, you may have an idea of which genes you want to test for significant correlations. In these cases, it is necessary to pre-compute the relative depths to which each cell has been sequenced as this calculation is based on all of the genes.
To do this, BigSur has a separate function: depth.scaling. This function will take in the original Seurat object and compute a vector of scaling coefficients for each cell to account for differential sequencing depth. This vector can then be passed to the main BigSur function in the depthlist parameter.
The following is some example code on how you might use this feature to analyze a smaller subset of data.
#Calculate the differential sequencing depths using the original seurat object
depths <- depth.scaling(example.seurat, assay="RNA", slot = "counts")
#Subset the seurat object to contain only genes of interest
genes.list <- c("eGFP", "ACTB", ...)
example.seurat.subset <- subset(example.seurat, genes.list)
#Run BigSur on the subsetted data with the depths vector as a parameter
output <- BigSur(example.seurat.subset, variable.features = F, correlations=T, depthlist = depths)
To change:
- Assert version specifics for dependencies
- Add more assertions to prevent improper use
- Change import structure to remove warnings from unused function conflicts
- Change minimum Fano factor from a set number to one calculated based on silhouette score.
- Add example correlation visualization using igraph
- Add custom igraph functions for visualization to package
In progress:
- Testing installation and code on Windows and Linux systems.