RNA/DNA binding site prediction on protein is a well-established field. Many computational prediction web service and programs have been proposed. A fair comparison of all the programs can improve our understanding of nucleic acid-protein binding mechanism. Currently existing prediction methods differ in:
  • 1) definition of binding sites;
  • 2) datasets used for training and test;
  • 3) stability of the prediction;
  • 4) metrics used in assessment;
  • 5) the ease-of-use.
  • With the help of results from this database, we get to know :1) the current state of binding site prediction programs; 2) the bottle-neck of the prediction problem; 3) the underlying principle of nucleic acid-protein binding.

    NBench (Nucleic Acid Binding prediction Benchmark)

         is a Database that contain necessary for benchmarking the current existing methods of Nucleic acid binding site prediction:
  • 1) links to all existing methods;
  • 2) all previously reported datasets;
  • 3) prediction result of all the methods on all the datasets;
  • 4) various assessment metrics applied on all the prediction methods on all the datasets.
  • Find the Results

    Besides, registered users can submit the datasets and prediction results of their own methods. Thus, we can better know how all of these new upcoming methods work. We strongly encourage submission of prediction results.
    Submit your predictions (Please register before your submission.)


    A large-scale assessment of nucleic acids binding site prediction programs.
    Miao Z, Westhof E. (2015)
    Plos Comp Biol 11(12): e1004639. doi: 10.1371/journal.pcbi.1004639.
    Supplementary Information:
    S12 Figures
    S15 Figures