Dihedral Angles Datasets for BenchMaking.



Aim of this page is to provide dihedral angle datsets, these datsets can be used to various purpose like development of new prediction method, benchmarking of existing methods.

We extract all PDB chains from ccPDB, these protein chains have number of residues between 50 to 3000, resolution better than 2Å, Rfree less than 0.25. Each ccPDB dataset was given a number (like ccPDB30) which represents redundancy level of the dataset. ccPDB30 means all the PDB chains in this dataset are not more that 30% identical (30% redundancy level). For more information on creation of ccPDB datasets kindly visit ccPDB: Compilation and Creation of Datasets from PDB


Download ccPDB30 dataset (4682 protein fasta files) --> Click Here

Download ccPDB40 dataset (5598 protein fasta files) --> Click Here

Download ccPDB70 dataset (7358 protein fasta files) --> Click Here

Download ccPDB90 dataset (8108 protein fasta files) --> Click Here

Download ccPDB95 dataset (8382 protein fasta files) --> Click Here

Download ccPDB100 dataset (12446 protein fasta files) --> Click Here