Sunday, August 11, 2024

Beyond protein lists! The possibilities of AI in downstream proteomics!

 


I put the figure in this post - not to make you not want to read it, but because I had trouble finding something funny to put at the top, but this really is a cool commentary in my post backlog


I recently had some of the worst "collaborators" of my entire career. You know the ones - where they want to drop off samples for you to prep, run and then stare at you with disbelief, contempt or hatred when you provide a curated list of proteins or metabolites and quantification values. "What do I do with this" they say or think or yell. Which always brings this to mind....

--- because, honestly, what was the goal in the first place? There is a whole permanent page over there somewhere with some of the next steps --->

And what you'll see is that most of those ideas for proteomics require putting protein lists into gene centric programs, and we know that's pretty silly. If you've actually taken a biology study all the way to the end yourself you've probably had 100 of your most differential molecules and typed them one at a time into the Google or the PubMed thing. What if instead of making Space Karen memes to anonymously post on Reddit you were using that to help find patterns in protein regulation data? 

Less funny, but probably a lot more productive! 

No comments:

Post a Comment