seurat icon indicating copy to clipboard operation
seurat copied to clipboard

Published 20 hours ago • verified publisher icon satijalab

Heatmap DEGs overlap: How to handle

Open Deepesh123gupta opened this issue 2 years ago • 2 comments

Here, I have so many clusters with DEGs but these genes are not very distinct for each clusters (to some extent which is normal). How to handle it? Data collected from scRNAseq. Suggestions are appreciated. image

Deepesh123gupta avatar Sep 16 '22 19:09 Deepesh123gupta

Hi, Could you please let me know what is each column representing? Is each column here representing a single cell or a pseudo-bulked cell type/cluster?

In addition, how is DEG defined for each cluster? Are they from FindMarkers()/FindAllMarkers() in Seurat?

longmanz avatar Sep 23 '22 19:09 longmanz

This is the average gene expression heatmap. Each column represents cluster. DEGs are defined using FINDAllmarkers() and plotted top 10 gene for 18 clusters (in seurat).

Deepesh123gupta avatar Sep 26 '22 13:09 Deepesh123gupta

Hello,

To find distinct genes for each cluster, you may have to run FindMarkers to compare similar clusters (if clusters are similar compared to the rest of your dataset, they may have the same markers that come up in FindAllMarkers, because this is a 1-vs-all comparison).

mhkowalski avatar Feb 10 '23 20:02 mhkowalski