I was reading a new paper about Alzheimer's. The paper is using a new technology called single cell sequencing, and was published in a prestigious journal Nature recently.
With the new technology, basically, researchers collect healthy and diseased brains, break the brains down to single cell-level, use new equipment to sort each nucleus, and quantify expressed gene copy number or copy number of particular genomic DNA sequence of interest in each of the nucleus.
Brain is not made of only neurons, but made of several different types of cells (excitory neuron, inhibitory neuron, astrocyte, microglia, endothelial cells, infiltrating macrophage, etc). Using known marker gene expression patterns as guidance, researchers can deduce type of the particular cell they are looking at.
The dataset is also publicized and stored in accessible archive in the journal. So we can look at the date in further detail.
That is all good.
One thing puzzling me is that the technology did not detect much brain aneuploidy (abnormalities in the genome) in another paper.
There is an old technology (called FISH) that looks at brain tissue on slides under microscope, then quantify brain aneuploidy by counting fluorescent signals. The old technology indicated high degree of aneuploidy in the brains.
Now, there is a discrepancy in results from the new and old technologies. In the case, which should we believe, or how do we reconcile?
There is no generalized rule. We need to decide on case-by-case basis.
At this moment, my judgement is inclined to side on results from old technology, with following reasons.
(a) Personally, I'd interpret that the new technology may be prone to cell isolation bias (meaning that aneuploid cells may not be isolated with current enzymatic process-method as well as other intact cells, thus representation error is occurring).
I have not checked the new dataset in full, so it is my speculation.
(b) Numbers of publication. There have been many papers with the old technology, while there are few papers with new technology yet.
(c) Simplicity of old technology. The old technology is much simpler; look and count. New technology is also sort of look-and-count by machine, but some black box process in it.
Jury is still out. In this case, both may be correct within the boundaries of the technologies.
Science makes progress like this. We need to keep an open mind.
The assay with new technology would cost us about $10-15K for a pilot analysis. I am working to get fund for the assay, so that I can see and compare the results myself.
PS
I am not a fan of the old technology. Aged brains (both healthy and diseased) accumulate auto-fluorescent materials, which gives big pain in using the fluorescence-based old technology. If we can use new technology to accurately observe brain aneuploidy, I'll switch to the new technology in a snap.
