I was looking at and comparing the miRNA target gene predictions from five commonly used algorithms, viz., miranda, targetscanS, pictar, microT and mirTarget. Surprisingly, there is so little overlap! And I also did a comparison with the entries in TarBase (that houses about 100 experimentally validated miRNA-gene pairs) and surprisingly almost all of the five prediction algorithms perform quite badly. For instance, none of these algorithms predict PTEN as a target of mir-21 (an experimentally validated mir-gene pair). Seems like an intersection of the 5 predictions is not going to help and we have to live with the union with some meta analysis. MAMI from Harvard does a decent job but even it misses several well known targets.