Data requirement for gene regulatory network

Question

[arunprasanna83]

Hello,

I have only RNASeq data for two conditions control and test (not a time series). Is it possible to identify gene regulatory networks ? If not, what are the minimum required datasets needed to construct one ? Kindly help me the way forward.

Thanks,

AP

Answer 1

[Marcel Ribeiro-Dantas]

Yes, you can infer gene regulatory networks with cross-sectional data, that is, measurements of samples at a specific point in time.

If you have a set of potential transcription factors, for example, you can try to measure the statistical association between all potential target genes and these potential transcription factors. This is sometimes called a TF-centric regulatory network and by only measuring the association between target genes and TFs you decrease the chance of false positive relationships (though increasing the likelihood of false negative relationships). There is much more to it, but a very superficial explanation is what I stated above.

With the regulatory network inferred, you could try to infer master regulators based on the differential expression analysis you can perform with your two conditions (control/test). All this can be done with the RTN R package.