Some example intensity data of MS/MS data.

My thought is that perhaps selecting local maxima may be a good way to cull the peaks in our PKL and DTA files. The basic algorithm: Select a peak if its intensity is greater than both the previous peak and the next peak.

Peaks that would be selected are hilighted:

No tests yet to see how this impacts the performance of MSMSFit or HMM_Score. (I am concerned that this method may prove to be too severe.)

Update: I’ve added a bit to the algorithm that adds a peak if there is a significant jump in mass from one peak to the next (e.g. 10 Daltons). Initial experiments are encouraging.