Methods to interrogate RNA-seq

Question

[telroyjatter]

Let's say you have an RNA-seq counts matrix for two experimental groups, and you have already generated your table of differentially expressed genes. Generally speaking, what are the best ways to interrogate this dataset?

The ones I can think of:

• Gene ontology enrichment
• WGCNA and/or MEGENA for co-expression networks: modules; eigengenes, eigengene networks
• ARACNE for transcription factor-target networks
• Key driver analysis (i.e., which hub gene in WGCNA/MEGENA/ARACNE has the highest connectivity; what other ways are there to define key drivers?)
• HOMER for regulatory binding sites

Any other methods, reviews, papers are highly appreciated!

[JC]

The first thing is "what is your biological question?" Every analysis depends on that response.

[telroyjatter]

Let's say the tissue is from brain, and we want to understand how the administration of a drug affects gene expression.

[Ram]

already generated ... table of differentially expressed genes

So you have a list of genes and are looking for biological meaning in this list?

[telroyjatter]

Or in the entire dataset, for example some methods include an entire counts matrix rather than just a list of DEGs or the counts of only DEGs.

[Friederike]

Echoing JC here: if you're only thinking of a question after having generated the data, things have been done backwards.

Why is the list of methods you've listed not sufficient to get you started?

[telroyjatter]

It is enough to get started. I don't even have data. I just want to create a list of all of the methods, with their respective applications, that can be used to mine, interrogate, and extract knowledge from an RNA-seq dataset (or *omics datasets more generally).

[telroyjatter]

If my biological question is to understand how a drug affects the transcriptome, are there tools or a particular workflow that would be recommended?

Answer 1

[Kevin Blighe]

Perhaps you can obtain some insight from my answer, here: What is the best way to combine machine learning algorithms for feature selection such as Variable importance in Random Forest with differential expression analysis?

Kevin

[telroyjatter]

Yes this is helpful. Thanks, Kevin!