CLIPSeed: Achieving High Precision miRNA Binding Sites Prediction using PAR-CLIP Data

MiRNA plays an important role in post-transcript regulation, many computational approaches have been developed for miRNA target prediction. Yet, the existing algorithms lack the capability to predict the true target when the perfect seed match presents in mRNA sequences and seed-match based methods still suffer from a high false positive rate. This paper develops a new machine learning based miRNA target prediction approach named “CLIPSeed”, which is a powerful, friendly, and flexible tool for predicting miRNA targets given mRNAs containing seed matches to miRNAs. Two versions of algorithms are provided based on whether the conservation features are considered or not, also user can specify FDR according to the practical needs. Given a miRNA and mRNA pair, CLIPSeed predicts whether this mRNA is a target of miRNA and outputs the binding sites information if it is a target. Five-fold cross validation on PAR-CLIP ALL-AGO (AGO1-4) data shows that CLIPSeed achieves >2 fold increase in prediction performance measuring by precision. Also, testing on independent CLIP data confirms good performance of CLIPSeed algorithm. Finally, human genome-wide testing using expression data by Gene Set Enrichment Analysis (GSEA) verifies that the predictions of CLIPSeed are significant and the predicted miRNAs targets are down-regulated at mRNA level. CLIPSeed achieves significant improvement in performance and highly reduces the false positive rate comparing to the existing approaches. Because CLIPSeed has high precision prediction, it provides biologists a more powerful tool for further study of cancer treatment and personalized medicine.