Monday, January 5, 2015
New study shows MudPit still has some big advantages!
Among the many things I've learned today? There is television show called MudPit. It appears to feature a mixture of live action and computer animated characters. Out of curiosity I watched 20 or 42 seconds of it. Worse things have happened to me. Off hand, I can't come up with many of them...
That aside, MudPit in Proteomics is an awesome thing. Online 2D fractionation that caused the coverage on our proteomics experiments to just sky-rocket, even on our older and slower ion trap systems. Recently, however, MudPit has fell out of favor with a lot of groups. For one reason, its kind of hard to set up. For another, you end up with an absolute ton of RAW data at the end and some programs are going to have trouble with it (Percolator, for example, in its 32-bit iteration can have trouble with that much RAW data at once).
Some labs still stick to it, and with good reason. Look at dozens of great papers that roll out of Mike Washburn's group at Stowers from the fleet of LTQs they have running MudPit around the clock. I've been lucky enough to spend some time there. The way they run it is really clever. Each sample is 24 hours so you can load the next one when you come in the next day. I'd argue that there has been as many pure biology papers roll out of that room full of LTQs as there has out of any other room full of instruments in the world.
So here is the question: If you have an ultra-modern nanoLCMS system, is there anything to gain from MudPit? Considering the amazing level of coverage that we see out of cell lysates in 1D separations these days, my initial guess would be no. I would have guessed that we only really gain a lot with MudPit when the instruments are slower or not as sensitive.
A recent paper from Aimee Rinas & Lisa Jones (IUPUI Lisa, not Fred Hutch Lisa) shows where this technology can help get low level peptides modifications that may not be detectable with one-dimensional separation alone.
In this paper (abstract here), they look at fast photochemical oxidation of proteins (F-POP) in a complex lysate on a Q Exactive. Even my favorite mass spec of all time has limitations. The F-POP modifications happen at exceedingly low stoichiometry and are tough to detect in a whole cell digest. By switching over to MudPit they an overall increase of 37% more modifications over 1D separation alone.
They go on to show that this relationship extends to less complex samples, by performing the same comparison on some huge protein complexes.
Can we learn stuff from going back to older methods? Absolutely, of course, ever time. It all depends on the experiment you need to do. If your freezer has a backlog of 2,000 samples in it, the added time and complexity for MudPit may not be worth it for you. But if you really need to get down in the grass for one super important sample, this technique is something you can keep in the toolkit. Of course you'll need to have 2 pumps/mixers on your LC to pull it off, but a lot of LCs are equipped in this way.
And please don't watch the Mudpit show...