search

broadinstitute / infercnv

Inferring CNV from Single-Cell RNA-Seq
Other
564 stars 166 forks source link

no obvious difference between immune cells and epithelial cells form tumor patients #574

Open zhangdae opened 1 year ago

zhangdae commented 1 year ago

Screenshot_2023-07-18-12-28-19-47_e39d2c7de19156b0683cd93e8735f348 I get this picture from infercnvpy,annotated cells by myself.as u see,no difference between ref cells and epi cells Screenshot_2023-07-18-12-28-47-31_99c04817c0de5652397fc8b56c3b3817 2nd picture is from infercnv,I cancelled the grouping of epi cells,still no difference. I performed sctransform to pre process data in seurat,and use Getassaydata(sce,assay ="RNA",slot="counts") to get raw data from seurat object. To verify my suspec，I randomly selected 15 samples，performed normalize,scale...then run infercnv，got a picture as below Screenshot_2023-07-14-06-29-04-38_99c04817c0de5652397fc8b56c3b3817 So,I wonder which way should I choose to preprocess seurat object before infercnv. SCT returned an assay named SCT,but there is no difference whether I use DefaultAssay ="SCT" or "RNA"

zhangdae commented 1 year ago

scrennshot really seems terrible，forgive me plz

cathy-y commented 1 year ago

I'm having the same issue, and my data makes it seem like something is wrong with the hierarchical clustering step. I'm running plot_cnv manually with cluster_by_groups and cluster_references set to False.

image
zhangdae commented 1 year ago

@cathy-y I set cluster_reference = T(Default = T), and now, trying to re-process the data,using Normalize,Scaledata...run infercnv once again.

zhangq220 commented 1 year ago

I'm having the same issue, and my data makes it seem like something is wrong with the hierarchical clustering step. I'm running plot_cnv manually with cluster_by_groups and cluster_references set to False.

image

Not only your data, I have the same issue.