How to verify putative miRNA

Question

[arcanjomjr]

My PI gave me a list of some DNA sequences, each one around 25 nucleotides, that are suspected to code for Vitis Vitinesfera miRNA. To know if they are complementary to some mRNA for a known protein sequence, we do the following process:

1. BLAST the sequence against a nucleotide database of Vitis Vinifera (TSA database)
2. Take some of the results and using their accession number BLASTX against a database of Vitis Vinifera (or related organisms) proteins (nr database).

Combining the results of 1 and 2, if any in both cases, we can know if the initial sequence is complementary to the mRNA of some known protein.

My PI wants me to automate this process, probably through writing a program.

This is my first contact with bioinformatics and is being of great learning value to me since I learned about Blast, Blast+, Biopython, etc. and I believe that I am arriving at a reasonable way to at least partially automate it.

Yet, this seems to be a problem so simple and so easy to come across that I find it hard to believe that there aren't any readily available solutions. I would like to ask if you know some program, or website, that does that or what would be the best way to arrive at such a solution.

Answer 1

[jens]

It looks like you are trying to perform miRNA targeting. Such short sequences are referred to as mature miRNAs. There is an abundance of tools performing miRNA targeting. For example: http://plantgrn.noble.org/psRNATarget/.

Additionally, you would want to confirm that the 25mer actually derives from a pre-miRNA. For that, there are also numerous tools. I would use https://www.researchgate.net/publication/337250440_Detection_of_pre-microRNA_with_Convolutional_Neural_Networks, but of course, I am invested in that tool.

If you want to use BLAST, try to first check what is available for your species in miRBase: http://www.mirbase.org/summary.shtml?org=vvi. Then go up in the taxonomy collect the mature miRNAs and blast them against your organism. At least this can show you whether the 25mers you received are already known to be miRNAs.

Sidenote: performing pre-miRNA prediction on eukaryotic genome-scale is difficult. Looking at all mature miRNAs on that scale without looking at pre-miRNAs seems a currently insurmountable task.

Verification

There is no computational verification. It needs to be performed experimentally for example using HITS-CLIP.

[arcanjomjr]

Dear jens,

Thank you very much for pointing out the "miRNA targeting" terminology, it lead me to discover a great deal about the subject, like Argonaute proteins, the seed region of miRNAs, its affinity for the 3'UTR of mRNA, the specificities of each organism and etcetera. Your article was also very informative and started me on learning about all this.

I came across a large number of tools for miRNA targeting, for example https://ccb-web.cs.uni-saarland.de/mtguide lists 98 of such tools, I didn't have success with even one of them, some websites are offline (like the psRNATarget you showed me), some aren't available for Vitis Vinifera, some didn't run on my operational system, some are web-based only and didn't accept big files, some simply didn't run on my computer, etcetera.

I still didn't solve my problem but learned a lot. I will talk to my PI for further instructions.

Thank you very much!

[jens]

I can confirm that psRNATarget is functional at the moment I am writing this (http://plantgrn.noble.org/psRNATarget/). There are also other tools RNA22 (https://cm.jefferson.edu/rna22/) or RNAHybrid (Vienna package: https://www.tbi.univie.ac.at/RNA/) which can help you achieve your task.

[jens]

The psRNATarget server moved: https://www.zhaolab.org/psRNATarget/

[boczniak767]

Yeah, miRNA could be only verified in wet-lab. Take a look at stem-loop PCR. I've used it and it works really well.

[jens]

Stem-loop PCR can identify shRNA. To verify that some RNA acts as a microRNA, one would need to show that the miRNA is bound to RISC and a target mRNA. That is what, for example, HITS-CLIP and PAR-CLIP can do for you.