This question may require complicated explanations:
1. I'm not sure if you set the neutral loss when setting up the modification, 
guess you have completed the setting. Because the mass of neutral loss is equal 
to the modification itself, this is also in line with your description of the 
modification label.
2. pFind 3 currently does not have the function of modification site 
positioning, so it will try to score all S and T, and find the highest. In this 
case, when the modification is added to the far left, the effect on the y ion is 
minimal (the y ion is generally relatively stable and contributes a lot to the 
scoring), which will make the score the highest.
3. As for the 1000 peptides without neutral loss, I'm not sure about this, 
because I don't know what the quality of the spectra are. In pFind, these 
spectra will not be automatically filtered based on the neutral loss.

Best,
Kaifei Wang

Greetings, I am a pfind user from Zhiyong's group in Carnegie Institution for 
Science. I am trying to use pfind to analyze o-glycosylation modifications of 
my proteins. O-glycosylation, such as O-GlcNAc which is included in my search 
parameter (HexNAc S/T), is usually labile and in my HCD data is expected to 
result in neutral loss and the exact positions of modifications are mostly not 
identifiable. I am surprised to find that all O-glycosylation have been 
assigned a modification site and there is no neutral loss. In the example in 
the attached pdf, HexNac is assigned to the first serine. The y/b ions do not 
seem to provide enough information to make this assignment, could you please 
explain to me how the site assignment is calculated? Also, over 1000 peptides 
are identified to be O-glycosylated, but none shows neutral loss, does pfind 
filter those out? Thank you in advance for your help.