hi, I am looking for differentially expressed genes by RNA-seq and used cufflinks for control and test samples.
cufflinks generated **gene.expr files and I assume these files can be used to compare the expression levels (FPKM values). Also cuffdiff generated direct comparison result with statistical p-values . According to the paper in Nat. biotect, the comparison result from cuffdiff shall be more accurate.
In my case, the 2 gene lists (from gene.expr comparison myself and from cuffdiff) are quite different, generally cuffdiff result shows more genes differentially expressed. After looking at the .sam files and coverage.wig files from Tophat, I found some genes genome locations are tiny different between control and test (especially genes with several transcripts) in gene.expr files, which leads to one zero values in cuffdiff results. cuffdiff possibly compares the control and test by gene location.
does this mean comparing gene.expr myself is a better solution here?
cheers
cufflinks generated **gene.expr files and I assume these files can be used to compare the expression levels (FPKM values). Also cuffdiff generated direct comparison result with statistical p-values . According to the paper in Nat. biotect, the comparison result from cuffdiff shall be more accurate.
In my case, the 2 gene lists (from gene.expr comparison myself and from cuffdiff) are quite different, generally cuffdiff result shows more genes differentially expressed. After looking at the .sam files and coverage.wig files from Tophat, I found some genes genome locations are tiny different between control and test (especially genes with several transcripts) in gene.expr files, which leads to one zero values in cuffdiff results. cuffdiff possibly compares the control and test by gene location.
does this mean comparing gene.expr myself is a better solution here?
cheers