RPmirDIP Predictions of ~6 Million miRNA-Gene Pairs

The RPmirDIP and RPmirDIP* predictors were used to re-score all available miRNA-mRNA pairs from the mirDIP database to identify novel putative interactions. The ~30 million pairs from the mirDIP database were re-scored using the RPmirDIP(*) method(s). Only those pairs with a Difference of Scores (DoS), defined as RPmirDIP(*)-mirDIP, with a value greater than 0.5 were published as these comprise the set of most likely candidate interactors for subsequent experimental validation. Of this set of ~6 million pairs, the data were divided into subsets based on a cutoff percentile (top 1%, top 5%, top 10%, top 33%, and bottom 66%). These data are presented in two sorted formats. The first is sorted on the Difference of Scores (DoS, RPmirDIP(*)-mirDIP) to identify pairs exhibiting the greatest augmentation of score over mirDIP due to the application of Reciprocal Perspective (RP). The second is sorted on the RPmirDIP(*) score to identify those pairs that RP is most confident in, regardless of the augmentation over mirDIP. Both formats support the exploration of potentially novel miRNA-mRNA interactions warranting subsequent wet laboratory validation. Note: To avoid data duplication, when selecting the top percentile of pairs (say top 10%), you must also select the top percentile files above that level (i.e. top 5% and top 1% files). For convenience, a single file containing all predictions is made available.