Image Source: 4designersart/Fotolia.com (lifted from this article)
Epidemiology has been one of those big things for a while. In my mind it seems like it kinda blew up the same time all this -omics stuff did. Schools have been putting lots of money into both the last few years. On the outside, it seems like they're opposite things. They are looking at trends in human beings to find disease patterns, while we're looking extremely deep into the disease, or people with the disease. However, now that we can get deep proteomic coverage in single runs, does this open us up to working together?
More and more, it looks like the answer is a resounding yes! For an overview of the topic you should check out this review: Epidemiologic Design and Analysis for Proteomic Studies: A Primer on -OmicTechnologies. It is open access and tries to bridge the gap.
I think it is very nicely written. While it is geared more toward the epidemiologists, in telling them what we do, it highlights some studies where the two were combined well. If I wanted to do a big study of some disease popping around in a human population, I wouldn't know how to sample people in a statistically valid sense. "Hey group A, you have the disease, right? You're TMT channels 126-129!"
Their job is to assess the factors that are important and design the experiment in a significant way. And then pull the right data out of the final protein list to show what is important. Turns out half the people here also suffer from a second disease? That's a nice data point to have so we don't draw a spurious conclusion, right? And there is something useful to be gained from that knowledge post-data processing? Even better!
This way we can focus on getting good quantitative protein IDs. And...if someone wants to explain what we do in terminology geared toward my collaborators' specific fields? Well! then I can send them this open source PDF to clear up some misconceptions before we sit down at the table and start designing this killer study!
No comments:
Post a Comment