The datasets are extracted from 47 protein-RNA complexes, which contains 107 hot spots and 102 energetically unimportant residues. A total of 79 hot spots and 72 non-hot spots from 32 protein-RNA complexes are randomly selected as the benchmark dataset and the rest 15 compounds are used as the independent dataset including 28 hot spots and 30 non-hot spots. The datasets used in PrabHot are available for download: Benchmark dataset Independent dataset