Seurat FindMarkers avg_logFC for integrated dataset

Question

[Payal]

Hi,

I am having trouble understanding the avg_logFC results of an integrated dataset (control and condition) in Seurat (3.1) for the FindMarkers pipeline.

gene_name p_val avg_logFC pct.1 pct.2 p_val_adj

gene1 0 -0.270265934 1 0.115 0

gene2 0 -0.332681178 1 0.167 0

gene3 7.25E-119 0.251635681 0.855 0.238 1.45E-115

gene4 2.10E-86 0.255847237 0.856 0.293 4.19E-83

gene5 2.84E-76 0.257140547 0.923 0.399 5.68E-73

pct.1 represents control in cluster1 and pct.2 represents condition in cluster 1.

According to Seurat vignettes - avg_logFC : log fold-chage of the average expression between the two groups. Positive values indicate that the feature is more highly expressed in the first group

What about negative values? For gene 1, avg_logFC is -0.2. Does that mean gene 1 is downregulated in control group in cluster 1? But why is percentage of cells expressed in gene1 more in control cells than condition cells?

Am I missing something?

Thanks

[crazyscientist]

Hello,

I am analyzing scRNAseq data by using seurat pipeline. I have two simple question. As per Satija lab tutorial, avg_logFC : log fold-chage of the average expression between the two groups. Positive values indicate that the feature is more highly expressed in the first group. pct.1 : The percentage of cells where the feature is detected in the first group pct.2 : The percentage of cells where the feature is detected in the second group

Then does the negative value of avg_logFC mean the feature is more highly expressed in the second group?

I have loaded samples such as 1st control and then 2nd treatment. by using Read10x function. In order find gene expression cluster19 between control and treatment, I ran the following code. So, in this analysis pct1 belongs to which group (treatment or control)?

cluster19.markers <- FindMarkers(Disease.combined, ident.1 = 19, min.pct = 0.25)

Could you please help me with this question? Thank you so much in advance.

Answer 1

[Payal]

I also posted this query in Seurat Github issues page. Please follow the conversation here - https://github.com/satijalab/seurat/issues/3397

Summary:

first switch to RNA assay

DefaultAssay(objecti)  <- "RNA"
object_RNA <- NormalizeData(object, verbose = FALSE)

then, find markers for every cluster compared to all remaining cells, report only the positive ones

object_RNA.markers <- FindAllMarkers(object_RNA, only.pos = TRUE, min.pct = 0.25, logfc.threshold = 0.25)