How to predict targets of miRNAs with indels in the seed region?

Question

[jhonatan.henrique]

I am currently doing scientific initiation, a type of academic research carried out by undergraduate students at Brazilian universities, in the area of bioinformatics. I am working with indels in the seed region of mature dog miRNAs. The goal of the research is to investigate which genes have become targets and which have stopped being targets due to the presence of indels in the seed region. However, the professor who is supervising my research and who I work with has only worked with SNPs in the seed region. She asked me to search for any literature on the prediction of targets for indels in the seed region of miRNAs.

After extensive research, I was unable to find any articles that performed target prediction for miRNAs with indels in the seed region. Could you give me some advice on how to deal with this situation?

Some of my data involve deletions, and some of these deletions remove the entire seed region of the miRNA. Considering that the region corresponding to the mature miRNA is part of a much larger sequence that constitutes the whole genome, should I consider the sequence of bases following the deletion region as the new constituents of the seed region and the miRNA? For example, in my data, the miRNA cfa-miR-8856 (5') corresponds to the sequence U[GAGUACG]GAGCUGGCUAGAGUGC (the bases within the brackets correspond to the seed region), and after the deletion, the miRNA became U[GAGUA]C. Should I consider the subsequent sequence as the new constituent of the miRNA, or should I consider the miRNA as lost?

Another example is the miRNA cfa-miR-30c, which originally was U[GUAAACA]UCCUACACUCUCAGCU, but had an insertion, resulting in U[GUAAAGC]UUCUGCUUCUAGGAACUGCUUGGAACUGUCUGGCUCACAACUUUCAUCCUACACUCUCAGCU. As you can see, there was an insertion of many bases. Should I still consider it a miRNA after all these base insertions, even though it now has more bases than a typical miRNA? If so, should I consider the whole sequence as a new miRNA, or just the amount of bases that a miRNA typically has, ignoring the rest?

Answer 1

[i.sudbery]

The first thing to check is whether the sequence will still generate a valid miRNA molecule. The sequence of the pre-miRNA most form a hairpin in order to be recognised by the miRNA processing machinery as an miRNA, if it doesn't then it will never be processed into an miRNA. If it does then it will still have to form a 19-22nt single stranded molecule to be function. The change in structure of the pre-miRNA could mean that its difficult to tell which 22nt will form the mature miRNA strand, but most likely, I'd guess it would be the sequence upstream of the 5' end of the miRNA (as Dicer/Drosha would now include this in the mature sequence).

[Buffo]

I agree with i.sudbery answer. The first step would be to test if those molecules can still form valid hairpin precursors to be processed by the dicer/drosha.