SSW: Seurat Standard Workflow (SSW)

In RuiyuRayWang/scWidgets: A collection of scRNAseq analysis functions

Description

Run Seurat Standard Workflow.

- Normalization
- Feature Selection (Highly variable genes, HVGs)
- Data scaling
- PCA
- Construct SNN graph
- Unsupervised Clustering
- Non-linear dimension reduction (tSNE, UMAP)

By default, we compute 50 PCs and use the 1~50 PCs as input dimension. To use an optimized dimensionality parameter, run JackStraw or ElbowPlot to determine the dimensionality of the dataset and supply to 'dims'.

Usage

SSW(
  object,
  assay = NULL,
  nfeatures = 2000,
  PC_features = NULL,
  npcs = 50,
  dims = 1:50,
  k.param = 20,
  algorithm = 1,
  resolution = 0.3,
  perplexity = 30,
  verbose = TRUE
)

Arguments

object	A Seurat object.
assay	"RNA" or "integrated".
nfeatures	Number of HVGs selected.
PC_features	If supplied, run PCA on these features.
npcs	Number of PCs to compute.
dims	Chosen dimensionality of the data. Used in SNN-graph construction and Non-linear dimension reduction.
k.param	Defines k for the k-nearest neighbor algorithm. Supplied to FindNeighbors.
algorithm	Algorithm for modularity optimization (1 = original Louvain algorithm; 2 = Louvain algorithm with multilevel refinement; 3 = SLM algorithm; 4 = Leiden algorithm). Leiden requires the leidenalg python. Supplied to Findclusters.
resolution	Value of the resolution parameter, use a value above (below) 1.0 if you want to obtain a larger (smaller) number of communities. Supplied to Findclusters.
perplexity	Perplexity used in RunTSNE.
verbose	Print output.

Examples

## pbmc_small is too small, must set perplexity to lower value
pbmc_small <- SSW(pbmc_small, perplexity = 10)

RuiyuRayWang/scWidgets documentation built on Dec. 18, 2021, 11:54 a.m.