Curiosity about DESeq2

Question

[tiago211287]

Taking a deeper look in the result of Differentially expressed genes I did find some strange thing. I show here the counts of genes for every group and the control. Look that group C and goup J has a negative fold change, indicating that this samples has less counts than the controls, and indeed it is. But, M has a slightly negative fold change, even having more counts in all samples, than controls. Why is it?

> gene  Control Control Control c   c   j   j   m   m   m   Log2foldchange c    Log2foldchange j    Log2foldchange m
> Txnip 2255    3417    3801    2103    1839    1926    2494    4051    5368    4765    -1.120.526  -1.061.433  -0.2599193

[Asaf]

Probably normalization