Skip to content

Step 2. Inferring Cell Type Factors using Seurat

This example demonstrates how to infer cell type factors from a cell-indexed SGE using Seurat.

This step starts with removing mitochondrial and hypothetical genes and filtering the hexagon-indexed SGE matrix by nFeature_RNA_cutoff. When X Y coordinate ranges are applied, the hexagon-indexed SGE matrix will also be filtered by coordinates.

Subsequently, this step applies sctransform normalization followed by dimensionality reduction through Principal Component Analysis (PCA) and Uniform Manifold Approximation and Projection (UMAP) embedding.

Next, the step employs FindClusters to segregate hexagons into clusters utilizing a shared nearest neighbor (SNN) modularity optimization-based clustering algorithm. During this process, FindClusters applies an argument of resolution to determine the "granularity" of clusters, i.e., a higher resolution value yields more clusters. A range of resolutions, including 0.25, 0.5, 0.75, 1, 1.25, 1.5, 1.75, and 2, are applied to explore the optimal resolution. For each resolution, the step generates an UMAP for dimensionality reduction, a spatial plot to visualize the clusters and their spatial arrangement, and a CSV file of differentially expressed genes for each cluster.

Additionally, this step generates a metadata file containing information on the cluster assignment for each hexagon, and an RDS (R Data Serialization) file that stores the complete Seurat object with all the compiled data.

Input & Output 

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
# Input: 
${output_dir}/${prefix}/barcodes.tsv.gz                                          # the cell-indexed SGE matrix from step1
${output_dir}/${prefix}/features.tsv.gz 
${output_dir}/${prefix}/matrix.mtx.gz

# Output: 
${output_dir}/${prefix}_cutoff${nFeature_RNA_cutoff}_metadata.csv                ## a metadata file
${output_dir}/${prefix}_cutoff${nFeature_RNA_cutoff}_SCT.RDS                     ## an RDS file
${output_dir}/${prefix}_cutoff${nFeature_RNA_cutoff}_res${res}_DE.csv            ## Each resolution returns a CSV file of differentially expressed genes for each cluster
${output_dir}/${prefix}_cutoff${nFeature_RNA_cutoff}_res${res}_DimSpatial.png    ## Each resolution returns an image of two panels including an UMAP for dimensionality reduction, a spatial plot to visualize the clusters and their spatial arrangement

Parameters:

  • --X_col: Specify which part of the hexagon ID corresponds to the X coordinate. For instance, in our example dataset, the hexagon ID is formatted as {X}_{Y}, i.e., the X coordinate is the first component. In this case, --X_col set this argument to 1. Default: 3.
  • --Y_col: Specify which part of the hexagon ID corresponds to the Y coordinate. As the Y coordinate is the second component in the example case, it should set to 2. Default: 4.
  • --nFeature_RNA_cutoff: Cutoff value for filtering hexagons by nFeature_RNA. Since this cell-indexed SGE is derived from histology files, nFeature_RNA_cutoff is set to be 0.

Commands: 

1
2
3
4
5
6
7
Rscript ${neda}/scripts/seurat_analysis.R \
    --input_dir ${output_dir}/${prefix} \
    --output_dir ${output_dir}/${prefix}/Seurat \
    --unit_id ${prefix} \
    --X_col 1 \
    --Y_col 2 \
    --nFeature_RNA_cutoff 0