RNA-seq analysis without replicates

Question

[fifty_fifty]

We have RNA-seq data for 12 samples for 12 conditions. Unfortunately, we do not have any replicates and each sample corresponds to one condition. For differential gene expression analysis, I will need at least 3 replicates (or patients) for each condition to be able to compare gene expression. I am interested in changes in particular genes across conditions. What can I do in this case? E.g., maybe there are some strategies for normalizing counts and comparing particular gene expressions.

Another thing I tried to solve this problem is to create artificial technical replicates using RESEQ. But apparently, the method is no longer maintained. I will appreciate any advice on this.

[rpolicastro]

I'm not sure if you have 12 samples representing 12 conditions that you can perform any meaningful analysis at the condition level. Is there any particular hypothesis you want to test that includes comparing one group of samples to another? Or perhaps there's some reason to try and cluster these samples into de novo groups?

[zau saa]

I think identifying context-specific gene after generating count matrix and filtering low-expression genes is the easiest way to accomplish your goal.

[ATpoint]

Appreciated that you're contributing, but it would be nice to suggest a "how" rather than a "what" since "context-specific" genes is basically what DEGs (so precisely what OP is asking for) are.

[zau saa]

Thanks for your reminder and sorry for my inaccurate reply.

To compare the three tissues and identify tissue-specific peaks, we used bedtools intersect -v (v. 2.29.2) for each pairwise contrast. By looking at the results of the multiple contrasts, we defined peaks exclusive to one tissue as tissue-specific. [1]

[1] Alexandre, P.A., Naval-Sánchez, M., Menzies, M. et al. Chromatin accessibility and regulatory vocabulary across indicine cattle tissues. Genome Biol 22, 273 (2021). https://doi.org/10.1186/s13059-021-02489-7

I followed the description above and that's what I really wanted to express.

Answer 1

[Gordon Smyth]

See "what do do if you don't have replicates" in the edgeR User's Guide.

Technical replicates, as produced by RESEQ, are of no use whatever.