No significant genes in RNA-seq analyses

[fahim]

Hi all
I am doing RNA-seq analyses using tophat and cufflinks. I followed all the steps of nature protocol papers without any problem except the 'cummRbund' portion.

I am adding selected snippet results below.

> 
cuff_data <- readCufflinks("diff_out")
> cuff_data
CuffSet instance with:
	 2 samples
	 35210 genes
	 65828 isoforms
	 46226 TSS
	 24611 CDS
	 35210 promoters
	 46226 splicing
	 20203 relCDS
> csVolcano(genes(cuff_data), 'C1', 'C2')
Warning message:
Removed 6924 rows containing missing values (geom_point). 
> gene.diff[gene.diff$significant=='yes',]
 [1] gene_id          sample_1         sample_2         status           value_1          value_2          log2_fold_change test_stat       
 [9] p_value          q_value          significant     
<0 rows> (or 0-length row.names)
> getSig(cuff_data,alpha=0.05,level='genes')
character(0)
> mySigGeneIds<-getSig(cuff_data,alpha=0.5,level='genes')
> getSig(cuff_data,alpha=0.5,level='genes')
character(0)
> length(getSig(cuff_data,alpha=0.9,level='genes'))
[1] 6759

I am amazed not to see anything differentially expressed. I am wondering if I did any mistake during the entire run.
My 'volcanoplot' also does not show any significant point in the label, however, I see some points that are above 2.0 on y-axis, demonstrating a p-value of 0.001.

I read some other posts suggesting to use edgeR.


Thanks and help appreciated