Question: In Seurat, How Do nCount_RNA Differ from nFeature_RNA?
1
gravatar for bioinformatics2020
10 months ago by
bioinformatics2020350 wrote:

I'm reading up on the Seurat user guide: https://satijalab.org/seurat/v3.1/pbmc3k_tutorial.html And they mention for QC utilizing

The number of unique genes detected in each cell and The total number of molecules detected within a cell

They then refer to them as nCount_RNA and nFeature_RNA, but I'm not sure which is which. So my question is:

1.) What are the nCount_RNA and what are the nFeature_FNA 2.) Later in the pipeline, when you're normalizing the data, it says they "normalizes the feature expression measurements for each cell by the total expression." Can anybody explain that?

seurat single-cell • 4.0k views
ADD COMMENTlink modified 10 months ago by jared.andrews076.9k • written 10 months ago by bioinformatics2020350
12
gravatar for jared.andrews07
10 months ago by
jared.andrews076.9k
Memphis, TN
jared.andrews076.9k wrote:

nFeature_RNA is the number of genes detected in each cell. nCount_RNA is the total number of molecules detected within a cell. Low nFeature_RNA for a cell indicates that it may be dead/dying or an empty droplet. High nCount_RNA and/or nFeature_RNA indicates that the "cell" may in fact be a doublet (or multiplet). In combination with %mitochondrial reads, removing outliers from these groups removes most doublets/dead cells/empty droplets, hence why filtering is a common pre-processing step.

The NormalizeData step is basically just ensuring expression values across cells are on a comparable scale. By default, it will divide counts for each gene by the total counts in the cell, multiply that value for each gene by the scale.factor (10,000 by default), and then natural log-transform them.

ADD COMMENTlink modified 9 months ago • written 10 months ago by jared.andrews076.9k

Hey Jared, appreciate the easy-to-understand response! Quick question, do you know how exactly Seurat determines the number of molecules within a cell?

ADD REPLYlink written 10 months ago by bioinformatics2020350

You know, I hunted around a bit and couldn't find exactly where nCount_RNA was defined. Presumably, it pulls that info during Read10x (from the .mtx file) or ReadAlevin and summarized for each cell to nCount_RNA during CreateSeuratObject.

ADD REPLYlink written 10 months ago by jared.andrews076.9k

Hi, Jared,

The definition of nCount_RNA (nCount_RNA is the total number of molecules detected within a cell) is pretty clear. However, I am still confused by the definition of nFeature_RNA (nFeature_RNA is the number of unique genes detected in each cell.)

What do you mean by "unique genes"? What are the detected genes unique relative to? For example, in a cell, there are 1000 genes detected totally and 400 genes detected uniquely. Does it suggest that these 400 genes are only detected in this cell? these 400 genes are not detected in any other cells? So these 400 genes are unique in this cell relative to all other cells? Am I correct?

Thanks.

ADD REPLYlink written 9 months ago by Research Dog0

Sorry, that phrasing was indeed a bit confusing - it's just the number of genes detected in each cell. They are not "unique" to that cell. I will edit my answer to clarify.

ADD REPLYlink written 9 months ago by jared.andrews076.9k
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 884 users visited in the last hour