RSEM output file format

Question

[mhasa006]

I followed the following tutorial for RNA-Seq data analysis. https://ycl6.gitbooks.io/rna-seq-data-analysis/

After performing the Differential expression analysis using R (https://ycl6.gitbooks.io/rna-seq-data-analysis/perform_de_analysis.html) the final output files are DE_analysis.gene.txt and DE_analysis_transcript.txt. The output file has the following format:

chr left    right   gid score   strand  name    source  biotype logFC   logCPM  F   PValue  FDR
1   14403   29570   ENSG00000227232 .   -   WASH7P  havana  unprocessed_pseudogene  2.890084925 2.977441296 103.3708945 8.25E-09    3.27E-08
1   89294   133723  ENSG00000238009 .   -   RP11-34P13.7    ensembl_havana  lincRNA 11.51392342 1.605598781 167.1675845 2.10E-10    1.21E-09

Could someone please explain the file format? Especially left, right, logFC, logCPM, and F. Which column represents the average gene expression value?

Answer 1

[Jake Warner]

Pretty difficult to evaluate exactly without seeing the contents of RunEdgeR.R but my guess is

chr = chromosome  
left = 5 ' start position on chromosome of feature 
right = 3 ' end position on chromosome of feature 
grid = feature code
score = ? 
name = feature name 
source = data source
biotype = feature type 
logFC = fold(2) change difference in expression(no idea what the test is)
logCPM = log(2?) counts per million 
F = ?
PValue = pvalue for logFC
FDR = false discovery rate for logFC

LogCPM seems to be the expression value you're looking for but I have no idea if it's averaged like you ask without the experimental design.

[mhasa006]

Thank you very much! I apologize for not being clear enough about the question, but your answer actually answered my question.