Overview

RSAPred is a web server to predict RNA-small molecule binding affinity using a combination of RNA sequence-based and small molecule-based features. RSAPred contains binding affinity prediction models for six classes of RNAs namely: Aptamers, MicroRNAs (miRNAs), Repeats, Ribosomal RNAs, Riboswitches and Viral RNAs (general and HIV TAR RNA-specific). The binding affinity values are predicted in log-scale (pK_d) for a given RNA sequence and small molecule representation. All models in the server exhibit an average correlation above 0.70 and a mean absolute error (MAE) below 1.0 during training, jack-knife (LOO-CV) tests, and on external blind test datasets. For more details on the datasets and model performance, please visit our statistics page.

The server accepts an RNA sequence and a small molecule SMILES (1D) or SDF file (3D structure) as inputs for prediction, along with the RNA class to which the sequence belongs. Users can check our tutorial page for further details on RNA classification and interpretation of results. If you have any further queries regarding the server or the models, please visit the FAQs page or write to us. Thank you for using RSAPred!