search

sulab-wmu / scPagwas

Obtain trait-relevant cell subpopulations by incorporating pathway activity transformed scRNA-seq data with GWAS data
Other
32 stars 9 forks source link

Input single-cell file processing #30

Open Gh-Genetics opened 4 months ago

Gh-Genetics commented 4 months ago

I am a single cell RDS file with multiple samples and batches. Should I use the integrated RDS file or the merged file. Because I found that when I use the following two treatments, the results may not be the same My single-cell RDS file: seurat An object of class Seurat 19283 features across 18818 samples within 2 assays Active assay: RNA (17283 features, 0 variable features) 1 other assay present: integrated 3 dimensional reductions calculated: pca, umap, tsne

The first type of processing: DefaultAssay(seurat)<- "RNA" Idents(seurat)<-"celltypes" seurat <- NormalizeData(seurat) seurat <- ScaleData(seurat, features = all.genes)

The second type of processing DefaultAssay(seurat)<- "integrated" Idents(seurat)<-"celltypes"

dengchunyu commented 3 months ago

In your dataset, there are two assays, with the emphasis placed on the standardized and batch-corrected assays, typically defaulting to the assay for RNA.