The benchmark dataset contains 404 protien sequences which composed of 178 cancerlectin and 226 non-cancerlectin, respectively.
The independent dataset contains 20 cancerlectin and 20 non-cancerlectin, respectively.
Close