I currently have a job as a research student working at Prince Labs at BYU, and so far it's been good. The funny thing is, though, that it's a Proteomics lab, and I know very little about that stuff; I've never even taken a chemistry class! But that's okay, because what I do is write Ruby programs to accommodate the Mass Spectrometry stuff. Aside from learning a lot more about Ruby and Linux, I've been able to see more how a CS degree can allow me to work in just about any field. I really don't need to know a whole lot about what the program is for, just how to get it to fulfill the required tasks.
What I'm currently working on is a Mass Spectrometry pipeline, which will allow users to execute a single command, and raw MS data will be transformed into identified proteins. My work can be viewed at Github.
I'm also working on an mzIdentML converter. My mentor (Is that a proper term for him?) thinks that if I make this converter then it will be something used by lots of people and maybe even get me mentioned in a paper. I have trouble believing such high claims, but I can at least say that no one is making a Ruby implementation, and with the power of Nokogiri, it's a clean and fast implementation. It would be cool though to create a program that would be useful to others.