I would suggest you to have a look to miRWalk database(http://mirwalk.uni-hd.de/). This resource offers possible miRNA binding sites within the complete sequence of known genes of humans, mouse and rats.
Moreover, it also has a comparative platform where you can generate a comprehensive overview of miRNA binding sites within mRNA 3'-UTR regions, resulting from 10 different prediction data sets. Previously conducted studies suggest that no program is superior to others, therefore a comprehensive atlas of the putative miRNA binding site predictions resulting from multiple algorithms are attracting more attention of the scientists who want to consider all the possible combinations (union or interaction) of available algorithms for their research. This database also has several new and unique features.