So I was struggling with this: Creating a dendrogram with a large dataset (20,000 by 20,000 gene-gene correlation matrix): Is there a way to use multiple processors (parallelize) to create a heatmap for a large dataset?
Now I think I found a good solution, taking a "meaningful" sample of the dataset, and then create a dendrogram-heatmap of the gene-gene correlation matrix generated from the sample.
I have got this far:
cluster3.seurat.obj <- CreateSeuratObject(counts = cluster3.raw.data, project = "cluster3", min.cells = 3, min.features = 200) cluster3.seurat.obj <- NormalizeData(cluster3.seurat.obj, normalization.method = "LogNormalize", scale.factor = 10000) cluster3.seurat.obj <- FindVariableFeatures(cluster3.seurat.obj, selection.method = "vst", nfeatures = 2000)
Now I am wondering, how do I extract a data frame or matrix of this Seurat object with the built in function or would I have to do it in a "homemade"-R-way?
I'm hoping it's something as simple as doing this:
cluster3.cells.variable.features <- as.matrix(GetAssayData(cluster3.seurat.obj, slot = "data")[, WhichCells(cluster3.seurat.obj)][, FetchData(cluster3.seurat.obj, var.features)])
I was playing around with it, but couldn't get it...
Any help would be appreciated.