"Missing proteins" are the ones that the genetics stuff says are there that we can't find proteomics evidence for.
Both the really hydrophilic and the really hydrophobic aren't a lot of fun to work with, particularly if you're using things like trap columns (
ProteoMiner has been around for a loooooooong time. And it had some solid applications in depleting proteins prior to doing 2D-gels. It's usefulness in more modern global proteomics has been a little less clear and you don't hear about it much anymore. The idea is that you make crazy big peptide ligand libraries and then use that as a depletion column. You know the top 10 depletion columns we use for plasma/serum? Imagine that on a much larger scale. Deplete 100 things?
This group uses it to go after the super hydrophobic membrane stuff. They start with a membrane prep then they ProteoMiner it -- then they use a combination of crazy HPLC separation methods and -- violin -- they end up with solid evidence for a bunch of proteins that should be there!
Are you looking for crazy membrane proteins? Maybe this is a cool new/old technique to check out!