After overexpressing or knocking down the miRNA (antagomir) evaluate the gene of interest by western. If the protein level remains the same regardless of the miRNA expression level than the miRNA does not regulate that particular target, at least under the conditions tested

Yes, I support Andrea Kasinski ,

If you can show that the expression of your gene is not at all dependent upon the concentration of the miRNA then it proves that the miRNA does not regulate that particular target.

and Ali,

you are asking for another model that would be negative dataset, but in this case that would not be necessary at all.

Dear Andrea & Debasish:

I meant ways to get this without experimental method. by means of computational or other methods.

Debasish: my project consists of "microRNA target prediction". to this I need 4types data as my algorithm input.

1) microRNA sequences

2) 3'UTRs sequences

3) experimental validated targets as posotive control data

4) negative data consisting microRNAs & genes that we are sure which they don't interact.

ok?

I suppose you are planning for a maschine learning approach

if you are going for computaitional method then you have to build your own negative data set.

1) microRNA sequences => from miRBase

2) 3'UTRs sequences => from UCSC

3) experimental validated targets as posotive control data => from miRecords/miRTarBase

4) negative data consisting microRNAs & genes that we are sure which they don't interact. => you can tweek the miR Tagget prediction softwares to generate this data set