Monitoring mutations in clinical samples is becoming routine...using genetics approaches. What if we could do it by mass spectrometry? That would be a pretty big deal. Why? Because protein is what does stuff! Mutations in DNA may not actually result in protein production due to missense and nonsense mutations that end up in proteins that are cleaved, or not even expressed (respectively).
Sure, I'm biased, but I want to see real proteins that are messed up. And it is hard to do. Proteomics requires databases in most cases.
Thats why this paper from Dasari et. al., out of the Mayo Clinic is so awesome. Here we see a 2-pronged attack on amyloid (no, I didn't know what an amyloid deposit was until I looked it up) mutations coming out of a certified clinical lab.
How'd they do it?
Like this! They started with a super fasta including sequences from known mutations. The MS/MS spectra were searched with the super fasta vs 3 search engines. The MS/MS spectra were also searched with Tag Recon and all the data was compiled. Boom! Extensive coverage of known mutations along with novel mutations pulled out from the super cool Tag Recon package.
Yes, this seems super elaborate. It is. Roman Zenka is listed as an author, and I feel like you can clearly see his hand in this workflow.
Most of us don't have the programming/bioinformatics resources the Mayo Clinic does, but this is a shining example of how deep we can search our data when we really really want to.
No comments:
Post a Comment