About gene expression data in FPKM stored in public databases

Question

[gspirito]

Hi everyone, here's my question:

Let's say that I want to know the expression level of a gene across different cell types. If I check online resources, such as Brain RNA-seq by Barres Lab (https://www.brainrnaseq.org/) I am able to get the average expression of my gene of interest in FPKM. For example:

I know that FPKM is not a good value to be compared between samples, let alone among different experiments. So how accurate are the results from this kind of online databases? Would that be an accepted result to put in a publication (as support for other results)?

Thanks in advance.

Answer 1

[Istvan Albert]

While the data may not be directly comparable, if you do recognize the limitations and avoid the obvious pitfalls you may be able to note behaviors that may be meaningful in some context.

If for example your gene of interest is suspected to play a role in the endothelium then that result above is an additional data point that is worth exploring.

What is best avoided is any type of large scale, brute-force search for patterns, the results found that way will be mostly artifacts.