A preprint describing our method for generating spectral libraries for the analysis of diaPASEF data is online. This method based is on using narrow window diaPASEF analysis split over several injections (i.e. gas phase fractionation) with coordiation in the ion mobility domain. TL;DR its better than libaries generated directly on the standard diaPASEF data and just takes 7 more injections of some representative sample. Maybe you can still eek out a bit more from a deep fractionation library but the signals you get are the least reproducibly detected. Parameter files that you load in timscontrol are in the supplement if you wanna give it a try.