Hi, I'm now conducting a ST analysis using Seurat R package. Normalization was conducted using SCTransform function and PCA was conducted.
I have a question about determining the number of PCs to be selected when conducting FindNeighbors and RunUMAP function.
When analyzing a single cell data, the number of significant PCs was determined by drawing an elbow plot and a jack straw plot, but JackStraw function using ST data that normalized by SCTransform function have an error. "JackStraw cannot be run on SCTranscrom-normalized data. Please supply a non-SCT assay."
I wonder how to judge the significant number of PCs in spatial data.