RPmirDIP Predictions of ~49 Million miRNA-Gene Pairs (Trained on mirDIP v4.1)

Brief Description The RPmirDIP predictor originally used to generate ~6 million miRNA-Gene pair predictions was retrained on the mirDIP v4.1 and was used to re-score all available miRNA-mRNA pairs from the mirDIP database to identify novel putative interactions. We release, here, all ~49 million predictions without filtering for RPmirDIP scores or the Difference of Scores (DoS) originally proposed. Example File Head miRNA,gene,mirDIP-score,mirDIP-class,RPmirDIP-score,RPmirDIP-class hsa-let-7a-2-3p,A1BG,0.0105377177207658,L,0.09713867,L hsa-let-7a-2-3p,A1CF,0.0641637887555465,M,0.45591232,M hsa-let-7a-2-3p,A2M,0.0204990488073111,M,0.16239062,L hsa-let-7a-2-3p,A2ML1,0.0132876238790541,L,0.1371971,L hsa-let-7a-2-3p,A4GALT,0.0385553499671089,M,0.30212826,L Example Python Code to Load into Pandas and Index by Columns For example, to load the data and search for a specific miRNA or gene, say hsa-let-7a-2-3p or A1BG, we can do the following in python: import pandas as pd tbl = pd.read_csv('all_pred_RPmirDIP.csv', index_col=['miRNA','gene']) tbl.xs('hsa-let-7a-2-3p',level=0) #index by miRNA tbl.xs('A1BG',level=1) #index by gene