Suggestions needed for mIRNA-human target prediction ?

Question

[k.kathirvel93]

Hi EveryOne,

I came up with the list of miRNAs after the deferential expression analysis of in-house data. Now i am trying to find human target genes for those miRNAs from miRDB, targetscan, miRWalk. Now i have a large list of Human target genes from these three databases. But i am confused, how can i filter the validated miRNAs from these list. Is there any easiest way ?. Thanks in advance.

[Kristoffer Vitting-Seerup]

I don't understand you problem? You have a list of DE miRNAs and you have a list with the predicted/validated targets for those miRNAs (from the DBs) - what is your problem?

[k.kathirvel93]

I have different number of Target genes from those three DBs(150 from miRDB, 205 from miRwalk), I donot need all the genes, i need some parameter to filter out meaningfully to comeup with best hits of target genes.

[Kristoffer Vitting-Seerup]

Then you might want to consider using anti-correlation across the entire Fantom 5 miRNA collection data (1800+ samples) which is described in this article.

Answer 1

[Kevin Blighe]

Hey,

Here is the way that I did it:

1, in silico predicted gene targets

In order to identify gene targets of micro RNAs (mirs), you can use miRNAtap (Pajak & Simpson, 2015). With this, you can choose different databases:

• DIANA (Vlachos et al., 2012)
• PicTar (Krek et al., 2015)
• TargetScan (Agarwal et al., 2015)
• miRanda (Betel et al., 2008)
• et cetera

Geometric mean score can be used to rank each gene. You can also choose to only include results that appear in 1, 2, 3, or more of te databases.

2, validated

With your output, search for these in the validated MicroRNA-gene Targets section of miRWalk2.0 database (Dweep & Gretz, 2015; Dweep et al., 2011).

Kevin

Answer 2

[zhxq]

you can use starBase database http://starbase.sysu.edu.cn/, it provides high-confidence miRNA-target interactions supported by CLIP-seq data.