Question: Discrepancy in DESeq2 fold change
gravatar for sophie.ehresmann
2.3 years ago by
sophie.ehresmann10 wrote:

Hello, I am currently analysing RNA-Seq data obtained from Ion Proton. I have triplicates of all my samples, for an empty vector, a WT protein, and 3 mutants. I am having trouble understanding the DESeq2 results. When I compare my WT and my empty vector, in the raw counts I have gigantic differences for my overexpressed protein (in the 60 000 range vs an average of 2 for the empty vector). However, after running the DESeq2 pipeline, the log2Fc is only .6, and it is not the highest FoldChange when it definitely should be. Also, I extracted normalized counts from the results and I transformed them using log2 then calculated the mean and sd and did the fold change, which comes to around to 6.7 . I understand it might be considered as an outlier and taken out of the analysis, but how can I force it keep these values? Secondly, when I manually calculate the log2 fold change in the same way (log2 of normalized counts, mean of log2, sd of log2, fold change of log2), I don't get the same results as when running DESeq2. Am I doing something wrong? Which result should I be looking at?

Thank you for your help.

rna-seq deseq2 R • 1.9k views
ADD COMMENTlink modified 2.3 years ago by Santosh Anand4.9k • written 2.3 years ago by sophie.ehresmann10
gravatar for Santosh Anand
2.3 years ago by
Santosh Anand4.9k
Santosh Anand4.9k wrote:

Deseq2 fold-changes are shrunken. See my earlier post: A: How to recover treated/control count from DESeq2 output

ADD COMMENTlink modified 2.3 years ago • written 2.3 years ago by Santosh Anand4.9k
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 509 users visited in the last hour