# Differential expression gene analysis

I have gene expression data in two samples like this:

Gene           s1   s2
C9orf152    2.96295 2.99861
RPS11       11.2968 11.0775
ELMO2       8.16786 7.6039
CREB3L1     6.24832 6.69416
PNMA1       8.11559 7.90189
MMP2        8.32177 7.06863
TMEM216     5.48133 5.68662
C10orf90    2.91206 3.11626
ZHX3        5.47916 5.31362
ERCC5       5.94527 5.92997
GPR98       3.09796 3.24092
RXFP3       5.17484 5.14564
CTAGE10P    2.71189 2.60001
APBB2       7.16389 7.40401


Now I want to know if any test can figure out the differential expressed genes between S1 and S2.

T-test can only give one p-value between two samples, right? I want to figure out each p-value for each gene between s1 and s2.

-
My answer (which I deleted) presupposed that for each gene you had multiple pairs of values so that you could do as many t tests as you have genes to compare s1 with s2. Then you can get an adjusted p-value by say a bootstrap adjustment for multiplicity or a Bonferroni bound or other multiplicity adjustment. But it appears that you only have one pair per gene. I don't think there is any test that you can apply with one pair per gene. –  Michael Chernick Sep 6 '12 at 23:31
Thanks, so if i add another sample S3 for each gene, the situation is the same? can't use test on that? –  user1586241 Sep 6 '12 at 23:41
For now, i can just use fold-change to compare the two sample. I just want to know if there are better statistical method can applied on it. –  user1586241 Sep 6 '12 at 23:42
It is not clear to me what you are doing . Can you explain what s1 and s2 and s3 represent? I thought you wanted to know if the gene expesses differently depending on the matter you are testing it with. To do statistical testing you have to not only compute a difference, you also need to know how variable that difference can be. –  Michael Chernick Sep 6 '12 at 23:56
s1, s2, s3 are different samples. Each sample have different gene expression level. –  user1586241 Sep 7 '12 at 0:21