search

satijalab / seurat

R toolkit for single cell genomics
http://www.satijalab.org/seurat
Other
2.2k stars 895 forks source link

Different cluster results with different orders of samples #1085

Closed mengchengyao closed 5 years ago

mengchengyao commented 5 years ago

Hi, thanks for developing Seurat. Great tool! I run seurat whith multiple_pancreas_workflow for 10xGenomics samples. I have a proble that we get different cluster results with different orders of samples and with the same settings. I could not understat, why?

Best Wishes

andrewwbutler commented 5 years ago

Hi,

Could you provide the code you ran that gave different results?

mengchengyao commented 5 years ago

Hi, I followed the multiple_pancreas_workflow pipeline. First run, commands are as follows:

ob.list <- list(sample1,sample3, sample4, sample2) pancreas.integrated <- RunMultiCCA(ob.list, genes.use = genes.use, num.ccs = 15) pancreas.integrated <- CalcVarExpRatio(object = pancreas.integrated, reduction.type = "pca",grouping.var = "tech", dims.use = 1:10) pancreas.integrated <- SubsetData(pancreas.integrated, subset.name = "var.ratio.pca",accept.low = 0.5) pancreas.integrated <- AlignSubspace(pancreas.integrated,reduction.type = "cca",grouping.var = "tech",dims.align = 1:10) pancreas.integrated <- FindClusters(pancreas.integrated, reduction.type = "cca.aligned",dims.use = 1:10, save.SNN = T,resolution = 0.4) I got 13 cluster for the result. Second run , when changed the order of samples,as follows. ob.list <- list(sample1,sample2,sample3, sample4) 16 clusters were get with the same settings.

The 4 samples are from 10xGenomics.

Best Wishes

timoast commented 5 years ago

Hi, sorry for the delay in response here but I will close this issue for now as we are no longer developing v2