Resolving single cell transcriptomics and proteomics data with pseudotime!

 

I have a very nice scSeq dataset (11,000 cells) that someone else generated that I've been looking at for years. I also have mediocre single cell proteomics (I did myself on much older technology) on the exact cell line and drug conditions, and it's cool when they seem to line up.

This new study attempts to resolve these differences with very smart biological models, and it's a lot of fun. 

I copied Figure 5 above completely because it's my favorite. 

What if you took a cell line and did something super extreme to it like inducing hypoxia (suffocating it!)? You could pretty much control exactly when the cells start panicking because of lack of oxygen, right? 

Under those conditions you'd think the transcript would accurately reflect what is happening. I mean, that's pretty genome critical. TIME TO RESPOND EVERYONE! Let's go! 

And...the...correlations are still sorta....meh.... The value in this dataset comes from the ability to dissect individual targets. Quick clarification, these aren't molecules taken out of the same individual cells. Some cells go to SCP and some go to scSeq. This deserved a lot more time and thought than I have, but I expect we can truly learn a lot from this study and these data.