I am trying to predict miRNAs against the gene of interest using online tools like Targetscan, miRanda ,miRsystem and DIANA micro T- CDS. As these tools produce a long list of miRNAs that might target mRNA, what factors should be considered while selecting candidate miRNAs for further study ?
Dear Bhoomika Sharma,
There is no reliable way to answer your question since all the prediction tools use different methods and differentially calculate prediction scores which makes impossible to correctly combine the results. But what you can do first is to set up a threshold on the prediction scores of each prediction tool in order to have most relevant miRNAs in your list. Second, you can use these miRNAs which prediction overlaps in different tools. But be careful with this because you may lose some important information. Third, if you are working with certain disease or certain condition of a biological system, try to check which miRNAs are differentially expressed in your model. These miRNAs should have a bigger impact on your model. Hope it helps a bit!
If I am correct you are interested in knowing the miRNAs that might target the mRNA of your genes of interest. I am not sure which species are talking about. If it is humans, there is nothing much to predict newly! Already people have used multitude algorithms to predict targets of miRNA that are known so far. All you need to do is look through the database for miRNAs that are predicted to target your gene of interest. Just find out of the extracted list, how many are experimentally validated. Its a good point to start. Start with those that have experimental proofs. Then move on to other miRNAs that are just predictions and not validated.
Dear Bhoomika Sharma,
There is no reliable way to answer your question since all the prediction tools use different methods and differentially calculate prediction scores which makes impossible to correctly combine the results. But what you can do first is to set up a threshold on the prediction scores of each prediction tool in order to have most relevant miRNAs in your list. Second, you can use these miRNAs which prediction overlaps in different tools. But be careful with this because you may lose some important information. Third, if you are working with certain disease or certain condition of a biological system, try to check which miRNAs are differentially expressed in your model. These miRNAs should have a bigger impact on your model. Hope it helps a bit!
Hi there
I completely agree with Simonas. Multiple predictors that simultaneously use several tools are the best approach for in silico target prediction. Overlaps between tools may give you some good clue to start.
Just be careful to compare predictions between tools, since all of them use different score parameters.
In this humble page you can find some information about multiple predictors, and which applications are used by them.
https://sites.google.com/site/mirnatools2/multiple-predictors
Best luck
Paco
Hi Bhoomika,
Good question!
Targetscan has not been updated since 2012, however, it gives two good scores: Probability of conserved targeting (Pct) and Total contextual score (TCS).
MiRanda has not been updated since 2010, but their mirSVR score considers empirical context to predict mRNA-miRNA regulation areas. MiRanda's algorithm is also used by MicroCosm (from EnsEMBL) database.
DIANA microT-CDS has the newest updated version (v5; 2013), their score is based on algorithms and bibliography extracts, they also indicate those cases that have been proved, and you can download your research in an Excel file.
As it has been said, it is very difficult to choose the best one. I recommend you to decide a criteria selection list: specie, databases (most recently updated as a reference, or in how many databases appears the target, ...), it's been corroborated (consulting PubMed papers), and/or which metabolic pathways and processes are you interested in...
Hope it'd helped!
Best luck
Elena
- Badges
- Science topic
- Similar topics
- Biological Science
- Biochemistry
- Biomolecules
- Nucleic Acids
- RNA
- Small RNA
- microRNA