I'm working on Human miRNA's, and I'm trying to predict miRNA targets in coding and non-coding genes. The algorithms(MiRanda, RNAhybrid, PITA) produce a lot of false positives and also there results differ significantly. So, now I use some online databases(TargetSCAN, starBase, miRtarbase, etc). However, I'm having some doubts while using these databases;
1) Do the miRNA's have different targeting genes in different cell lines(HUVEC, HELA, etc)? 2) Does the miRNA's have an opposite effect on its targets in Tumor and normal cells?