both log2 and log2fc giving a completely huge difference of a gene in subpopulations
1
0
Entering edit mode
10 months ago
am-cell • 0

Hello, I am very new in the field of bioinformatics and data analysis, and I have a basic question and I would appreciate your help. Why find differences in the value of a specific gene in some subpopulations using the log2 mean expression distribution and log2FC differential expression output, both log2 and log2fc giving a completely huge difference? this gene is expressed in a specific population x using the log2 mean, and by using the log2fc, it shows in another population more compared to the log2! is that make sense?

loupe-browser single-cell-sequencing • 966 views
ADD COMMENT
1
Entering edit mode

Please add any data, plots or anything to illustrate the problem.

ADD REPLY
0
Entering edit mode

![enter image description here][1]

enter image description here

ADD REPLY
0
Entering edit mode

Yes. asking why expression levels and fold changes are different. I did t expect this difference, and which one to consider? for info the other genes didn't have this issue of this huge difference just some genes.. thanks

ADD REPLY
0
Entering edit mode
10 months ago
LChart 4.6k

Because in loupe the logFC is not computed with respect to the same set of cells: it is always X vs not X. What that means is, if (in this case) population a and population d express a gene, and all other populations do not (or do not express it well), and d is much larger than a, then mean(d) - mean(not d) is much larger than mean(a) - mean(not a). And if z is a small number of cells, then mean(z) - mean(not z) will be the smallest, as not z contains both a and d.

I suspect that's what's happening here.

ADD COMMENT
0
Entering edit mode

Thanks for the answer. is this mean that the gene is differentially more expressed in the population d compared to other populations? even though looking to the log2 mean expression that shows clearly the gene expression is more in z population! I get confused which graph to consider and why this differences!

ADD REPLY
0
Entering edit mode

another question, is it correct to transform these graphs into percentage %?

ADD REPLY
0
Entering edit mode

yes z has smallest number of cell, however d is not larger but it is the a population.. in term of cell number!!

ADD REPLY

Login before adding your answer.

Traffic: 1833 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6