Chimera: An Open Source RNA Pipeline for Long Covid Research



“
I spent about 200 hours over the last two months building an RNA pipeline to analyze my own long covid blood RNA sample. This post talks about my efforts and the end result. While I plan to open source some of this at some point, until I get more data, I'm unable to refine it enough to release it so it can eventually be used by researchers. But thanks to some new relationships with a few labs I've recently established, one of which is pretty enthusiastic, I should be able solve that problem soon.

So here is a summary of some of my efforts for creating my own RNA pipeline to help further #longcovid research, along with what it showed for my own blood sample.”
 
That's an astonishing read. Especially the stuff about his b-cells.

Individual efforts like this could really help with hypothesis generation for big studies.
 
Murph said:
That's an astonishing read. Especially the stuff about his b-cells.

Individual efforts like this could really help with hypothesis generation for big studies.
Click to expand...
Props to this person for setting up the pipelines on their own. This kind of analysis is basically what I do every day. Unfortunately I have to rain on the parade about the usefulness of this data. Getting data “calibrated” to a control set sets off every alarm bell in my head [edit: esp if the control is collapsed and you don’t see each sample individually]. And the interpersonal variability in an RNA-seq dataset is usually very high—individually it makes it nearly impossible to figure out which genes are relevant to any phenotypic trait.

They’re also claiming that they have clonal expansion of B cells but their screenshot says TCR which is for T cells. It lists IG genes so I’m guessing it was BCR and the AI summary messed up. It also says it’s 18% of all IGH—not of all clones. I would be very skeptical of calling that evidence of clonal expansion nor antigen persistence.

I hate to be the downer but there is a reason people spend years becoming experts in these types of analyses. Without understanding of both the steps in the pipeline and the biological context it’s unbelievably easy to walk away with nonsense interpretation. Add AI into the mix and it becomes word salad masquerading as biological insight
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom