Is it possible that I have anomalous and LLG peak, but I have no difference density peak? In this case, is that because the model I have is not good enough or the diffraction data at this site of the model is missing?
You should at a minimum see > 1sigma density in the 2mFo-DFc map, and if the site is unmodeled (or modeled as a water) you should see an mFo-DFc peak as well. If neither of these applies, the anomalous and LLG peaks are probably just noise.