res.ape <- lfcShrink(dds=dds, coef=2, type="apeglm")

For the apeglm option, please is the coef=2 a constant value and is it determine by any other thing especially when my count data contains 3 rep for each samples. Thanks

The tool is called DESeq2, not DSEQ btw. No, coef corresponds to the output of resultsNames(dds) indicating the respective comparison (the coefficient) you are interested in. If you do not find the respective coef in resultsNames(dds) then you will need to relevel the groups, see https://www.biostars.org/p/448959/#484944