RNApred is a webserver developed for predicting RNA-binding proteins from their amino acid sequence using different compositional features of proteins. In addition one module of RNApred also uses the RNA-binding residues predicted by the RNA-binding amino acid prdiction method PPRINT.
Datasets: The prediction method was developed using a dataset of 377 RNA-binding and 377 non RNA-binding proteins. These proteins were obtained from Swiss-Prot
Prediction Approaches:
- Amino Acid Composition Based: This method has been developed using amino acid composition of proteins.
- PSSM Based: This method has been developed using evolutionary information in form of PSSM profile obtained from PSI-BLAST search against NR protein database.
- Hybrid Approach: This method has been developed by adding the RNA-binding amino acids prediction method PPRINT to the PSSM based approach described above.
RNApred allows user to submit more than one sequence for predicting RNA-biding proteins using composition based SVM model. In case of PSSM and Hybrid approaches, only one sequence at a time will be accepted. If user submit more than one sequence, only first sequence will be considered for prediction.
The probability of correct prediction directly depends on the threshold. For prediction with high confidence (less probability of false positive prediction) high threshold should be choosen.
