miRNA target predicition - Too many targets come out... (Dec/05/2007 )
I am looking for the target of a specific microRNA. My problem is there are too many predicted targets there, like Sanger database (by miprovides over 800 targets, and TargetSan provides over 100 targets. What shall I start with? Just pick a whole bunch and clone the 3'UTR, and keep trying one batch after another? This sounds ridiculous!!
What is more, the target lists from the two database hardly overlap with each other...This makes me even less willing to start the fishing game.
Someone have any suggestions? Thanks very much.
This is the dilemma of the miRNA field: so many targets predicted and few are true. I would start with those that are conserved across different species and have high quality scores (not just seed sequence matches). By doing so, I have recently identified two true targets on a gene. These two targets are not predicted by any program.
not sure if you've used it, but have you tried the pictar algorithim?
http://pictar.bio.nyu.edu/
i remember reading about it nature, and it seems ok.
V
Thanks. It is so great that you have identified true targets for your miRNA!
Do I understand things right that your true targets are not predicted by any program? In that case, how did these two come across to you?
Yeah...That's a nice one. The problem is the particular miRNA I am studying is not on the list of PicTar's:-(
http://pictar.bio.nyu.edu/
i remember reading about it nature, and it seems ok.
V
Do I understand things right that your true targets are not predicted by any program? In that case, how did these two come across to you?
Hi Yuut,
I downloaded the 3UTR sequences of my gene for both human and mouse and also miRNA sequences. I ran my own target prediction using the miRanda program. the miRNAs should also be conserved among species because there are some miRNAs for example for human that are not conserved and should not be trusted.
luckily enough, the targets I found work in both human and mouse cells.
One option for target confirmation is to block the target with a Morpholino oligo. No cloning needed.
Choi WY, Giraldez AJ, Schier AF. Target Protectors Reveal Dampening and Balancing of Nodal Agonist and Antagonist by miR-430. Science. 2007 Oct 12;318(5848):271-4. Epub 2007 Aug 30.