Improve your false discovery in your match between runs with PIP-ECHO!

For an old and probably inaccurate description of match between runs (MBR) you can check out this old post. 

Also, you probably shouldn't go past Fengchao and Sarah's paper here. Link might be the preprint.

Quick breakdown, though -

Imagine you run 50 LCMS runs on different patient samples.

In 35 of those runs you fragment and successfully identify PEPTIIIIDEK, it's pretty much 100% +2 charged and 634.3608 and comes off at 15.6 minutes 

In the other 15 runs you see a +2 peptide at 15.6 minutes but you don't fragment it or don't get good enough sequence quality for a positive ID. 

Match Between Runs (MBR) to the rescue! It donates that identification to the runs where it was not identified. 

Perfect idea, right? What's the problem? There are crapload of peptides in any tryptic digest and they coelute a lot. And as the dynamic range of our instruments keeps going up we see lower abundance peptides that we might not have before. 

Compound this with shorter LC gradients

And the fact that every mass analyzer has a +/- mass error

And the retention time on nanoLC, which everyone is pretty much using for some reason no one can justify, is probably more accurately, in those 35 runs, that peptide is coming off somewhere between 14.5 and 16.5 minutes - and now you might be quantifying the wrong peptide. 

The top link above is to IonQuant which works in FragPipe. 

Could you take that idea and build something even better? Maybe! Just do this! 

The comparisons look good, though! And there is some serious nerd power on the preprint. 

I checked the updates and it doesn't look like it's live on MetaMorpheus yet, though something that was driving me absolutely crazy a while back is! (Thank you! I thought my brain had broken (the diagnostic ion is wrong a lot of places - including in software from my group. ) 

Knowing these people we'll PIP-ECHO it in one of the upcoming builds. Fingers crossed it will work outside of the Orbi domain!