Deep Visual Proteomics has been a major topic of most meetings I've been to for the last 2 years or more and even made some real life headlines. However, if you really dug into the first paper you might have ended up where I did thinking "wait. wtf? is this just laser capture microdissection with some fancy informatics and very large assumptions and extremely good marketing pulling it all together? The actual reasolution of the first paper came in at around 100 cell size cuts and because cells stained with a marker or two they were grouped together. So it was a lot more like the "single cell type" thing you can do by FACs sorting 100 cells from a population based on the expression of a cell surface molecule, but cut across a tissue. Still cool....I mean...ish...but still cool....
THIS one is cool.
In paper #2, the resolution of the cuts goes way way way up. Based on my recent crash course in how laser capture microdissection works, I think that this means the speed goes way way way down. However, the size of the pieces coming out of this are about the size of a hepatocyte (300 - 700 picograms of protein depending on the cell and conditions). The cuts are somewhere in the 1/2 that size. HUGE step forward.
I guess my trepidation with the first paper is the intrinsic assumption that sufficient expression of the same cell surface proteins means that these cells are the same. The value is there, though, in the though -- "hey! these cells across this slice are clearly different than these over there - Let's dig deep into their proteomes and see why they're special." By cranking up the resolution like this, even on a few cells you can test that assumption that Marker A is a definition of a cell type!
No comments:
Post a Comment