Benchmarking SILAC workflows in 2025! Wait. What?

Okay - for just a second I thought I'd mistakenly scheduled posting this article about 10 years in the future, but apparently this really is new! 

Mandatory - 

For any of you scientists out there who aren't getting up in the morning complaining about your joints, SILAC was something we used to do in proteomics all the time. We did it to the point it was called the "gold standard for proteomics quantification" all the time. And not just from the companies that sold us the heavy labeled amino acids which caused every plate of cells to cost $183,000. And at this time that was a big deal because to do good comprehensive proteomics on an Orbitrap XL in 2009 when I was doing it required 15 plates of cells. If you did a great job, 3 weeks of run time would get you 3,000 proteins quantified. Please not that some of these numbers are exaggerated.

Anyway, you'd grow cells passage after passage in media with heavy lysine and arginine until you were pretty sure that all your proteins had heavy labels. Then you'd pretend that you didn't think that cells were way too dumb for something like 18 passages in very strange isotopically labeled amino acids could have any possible phenotypic effects. Then you'd take cells grown without it, treat one with drug and one with DMSO, then pretend that DMSO has no phenotypic effects. Then you'd lyse your cells and mix your light and heavy proteins or digested peptides (I forget, I last used it for a paper in 2010? 2009?) and run them. At the MS1 level you'd see your heavy/light pairs and quantify off those.

There were drawbacks, some of which could probably be inferred from my description of the method above, but a lot? at least some? good science came from it. I can't think of any off the top of my head, but you've probably heard my philosophy that it's best to ignore everything in proteomics before 2017 -and this technique was largely gone by then. 

However - if you did have some reason to do SILAC in 2025 - I bet you'd wonder what could and should process the data! And here you go! 

Silliness aside, I've never considered doing SILAC DIA. 

Oh yeah, you can do some really cool stuff with SILAC by introducing it and then changing the media. That can provide measurements of protein half-life and protein turnover and things like that. There are reasons. Just don't use it for pairwise drug treatment stuff. There are much better ways to do those things now!