Sorry for this silly question. I use rlog trasfrmation and I found many difference from my samples. There is many values are negative.

rld = rlog(dds1)

apply(assay(rld),2,summary)

Min.        0.6044    0.02705   -3.248   -3.063   -2.740   -2.165     -3.181   -3.004
1st Qu.     2.5560    1.75300    4.208    4.335    4.391    3.567      4.165    4.133
Median      3.2620    2.40200    6.626    6.546    6.387    6.328      6.500    6.496
Mean        4.4300    3.53900    6.201    6.149    6.096    5.746      6.180    6.054
3rd Qu.     6.1900    4.46900    8.467    8.297    8.121    8.383      8.443    8.346
Max.       17.5900   17.51000   17.130   16.410   16.680   16.270     17.620   17.500

Is it normal obtain negative values after trasormation?

Here's one of the posts I found by searching "rlog negative":

A: Alpha diversity on Rlogtransformed data