From a user and instructor of Skyline:
"When you correlate two spectrums with the mirror option in Skyline a dotp is calculated, does it take into account the unfragmented precursor peak? I ask this question because I have noticed that PROSIT and other prediction tools (like pdeep3) do not include the unfragmented precursor intensity in the prediction and I guess this affects negatively in the dotp value. Am I right? I attach you some examples in a word document."
She also referenced this paper as a useful source:
Whe should either fix what we do for dotp in this case, or add an option, if we feel what we are currently doing has some utility.