Seurat3 DimPlot inconsistency

0

Entering edit mode

3.8 years ago

mbk0asis ▴ 680

Hi, all!

I'm currently analysing a single cell RNA-seq data using Seurat3.

While analysing, I found that the UMAP/tSNE plots by DimPlot kept changing everytime I ran it even though I sat the seed.use for RunPCA and RunUMAP.

The code I used is below.

>nDim = 50
>mipsc <- RunPCA(mipsc, npcs = nDim, features = VariableFeatures(object = mipsc), seed.use = 123)
>ElbowPlot(mipsc, ndims = nDim)
>mipsc <- RunUMAP(mipsc, dims = 1:20, seed.use = 123) 
>DimPlot(mipsc)

Is there a bug in the code or other steps I must set a seed?

Thank you!

RNA-Seq Seurat Seurat3 • 1.8k views

ADD COMMENT • link 3.8 years ago by mbk0asis ▴ 680

0

Entering edit mode

Hi,

My version of Seurat is 3.1.5. In the docs of runPCA() and runUMAP says: seed.use: Set a random seed. By default, sets the seed to 42. Setting NULL will not set a seed.. By default is using the seed 42, so there is no need to set the seed at all in first place (I mean, if you just want to keep reproducibility).

This seems odd to me. Therefore, you might consider open an issue on their github repo: https://github.com/satijalab/seurat/issues

António

ADD REPLY • link 3.8 years ago by antonioggsousa 3.2k

0

Entering edit mode

Thank you, António!

I will do that~~

ADD REPLY • link 3.8 years ago by mbk0asis ▴ 680

Login before adding your answer.