1 of 1000 things I love about cryosparc is dragging duplicate particles from several series of classification into a new job with each other, knowing the duplicates are removed automatically. Run classification 4 times with different parameters, take the desired class(es) from each run and pile them together with ease. Moving forward you have every particle that ever contributed to that desired volume.
It would be awesome to be able to do a similar task, where we can take all particles from multiple inputs, and keep only those which are common. Using the above analogy, we would keep only the particles which contributed to the desired class EVERY time. Every input must have those particles or they aren’t used. An easy way to guarantee that the particles moving forward weren’t incompletely classified out, but were constantly a contributor to the desired subset, for instance.