Reproduce clusters

[NicolaasVanRenne]

Dear all,

Is there any way to reproduce the clusters in the paper (MacParland et al, Nat comm), starting out from the UMI file on GSE database? (GSE115469_Data.csv)

If I make an object GSE115469 out of the csv file, and run the following I get a Tsne plot, with 18 clusters that looks nothing like the one in the paper... Any hints?

eb1S <- CreateSeuratObject(exprs(GSE115469),"e15",min.cells=0,min.features=0) eb1S <- FindVariableFeatures(object = eb1S) eb1S <- ScaleData(object = eb1S) eb1S <- RunPCA(object = eb1S) eb1S <- FindNeighbors(object = eb1S) eb1S <- FindClusters(object = eb1S) eb1S <- RunTSNE(object = eb1S) DimPlot(object = eb1S, reduction = "tsne")

[JulieBaker1]

This project don't follow the Seurat pipeline. You can view the source code.

[NicolaasVanRenne]

I know - this was april 2019 and my early days in scRNAseq analysis :)

You can ask the original authors for the R object of their analysis. That way you can load it straight into R and explore with seurat toolbox. Ask Brendan Innis, he is very helpful.

Best, Nicolaas