I work with datamining to found potential target genes. I advise you try with different algorithms for higher posibility to find them, for example TargetScan, Mirwalk, miRbase, etc. There are many database. The choice of these database dependent on that combination of settings you decide consider (conservation, folding, energy for binding, binding sites, score of prediction). Please check this article when describe one of the methods in silico that I have performed.
Sylamer and more precisely SylArray developed by EMBL http://www.nature.com/nmeth/journal/v5/n12/full/nmeth.1267.html https://www.ncbi.nlm.nih.gov/pubmed?holding=&db=PubMed&cmd=search&term=20871108 might be a good option for you. While attending miRNA course by EMBO I found this method is very well defined and better than any other and still believe.
Another aproach could be trying to predict the targets of the miRNAs you are interested using a trained model on previous expression data (targetExpress). If you check the background of the paper, you'll find references to other software taking advantage of available expression datasets with and without miRNA misexpression.
http://targetexpress.ceiabreulab.org/
Article Improving microRNA target prediction with gene expression profiles