Dear All,

I am working on RNAseq data. I am comparing tumor and normal samples. I got this output file from cuffdiff step.

I would like to understand about this output file.

gene_exp.diff - Gene level differential expression. Tests differences in the summed FPKM of transcripts sharing each gene_id.

If you have any document discussing the below things, kindly refer it. Thanks in advance.

I have following queries,

• value1 is FPKMx and value2 is FPKMy

• Log2(fold-change) => log((6.04054/158.412),2) is -4.71286 for gene MUC1. Refer row 6 in the below image.

- How to calculate test-stat by comparing one value from each sample?

- What is the threshold value used to categorize genes as "significant" and "non-significant"?

• For example, I noticed the following fold change for the below genes are referred as SIGNIFICANT

**DMBX1 : 0/3.51818 is -infinite fold change

SPRR2F : 45.7115/0 is infinite fold change MUC1 : 6.040/158.412 ( sample1 is 25 times higher than sample 2)

ADAM12 : 0.592487/100.418 ( sample1 is ~100 times more than sample2)

VEGFA : 2.49672/57.0407 ( sample1 is ~27 times more than sample2)

ADGRG2 : 0.119485/8.46 (sample1 is ~7 times more than sample2)**

• For example, I noticed the following fold change for the below genes are referred as NON-SIGNIFICANT

**AGRN1 : 6.30387/57.554 (sample1 is ~9times higher than sample2)

MMP23B : 22.1959/9.17679 (sample1 is ~2.5times lower than sample2)

LINCO1128: 3.11434/1.87709 (sample1 is ~1.5 times lower than sample2)**

https://i.imgsafe.org/190431586c.png