miRNA-mRNA interaction

[ChrisK]

Hello,

I am currently searching for methods to determine miRNA-mRNA interaction pairs based on (RNA-Seq) read count data. Following the great review of Muniategui et al. 2013, one main approach is to determine the correlation between miRNA and mRNA expression over several samples. Alternatives are regression-based or even Bayesian, and advanced methods include normalization of reads, restrict to differentially expressed observations, or test only pairs infered with computational target prediction.

This is already a lot of information, but I want to ask the community about practical experiences with these tools. Can you recommend a particular approach or tool (esp. for human cancer data)? What are your experiences with straightforward approaches, e.g. a simple (e.g.) Pearsons correlation coefficient on normalized read count data?

Then I've also found tools like Corna, Sigterms and MMIA, that use enrichment in gene lists, but, to my understanding, do not directly report interactions. Does anyone have experiences with these tools or can you recommend similar packages?

Thanks in advance for your comments!

[TiborNagy]

We have worked with a non-model plant organism. We made a differential expression analysis (Trinity) and a miRNA expression analysis (UEA) and we used Pearson correlation. We also used degradome analysis to confirm the interactions.

Without degradome we get too many potential interacions which is suggest me a large number of false positives.