ccPDB2.0: an updated version of datasets created and compiled from Protein Data Bank

ccPDB 2.0 is an updated and significantly enhanced version of the database of datasets created and compiled from the Protein Data Bank (PDB). This resource provides researchers with high-quality, non-redundant datasets of various protein properties (such as ligand binding, secondary structure, and metal interactions), serving as a gold standard for training and benchmarking machine learning models in structural biology. Web Server: https://webs.iiitd.edu.in/raghava/ccpdb/ Citation Agrawal, P., Patiyal, S., Kumar, R., Kumar, V., Singh, H., Raghav, P. K., & Raghava, G. P. S. (2019). ccPDB 2.0: an updated version of datasets created and compiled from Protein Data Bank. Database, 2019, bay142. https://doi.org/10.1093/database/bay142 About the Research The exponential growth of the Protein Data Bank (PDB) makes it challenging for researchers to manually curate clean, non-redundant datasets for specific structural studies. ccPDB 2.0 automates this process by systematically categorizing protein structures based on their interactions and structural features. Expansion: This version includes 41 different types of datasets, nearly double the amount provided in the initial version. Non-Redundancy: All datasets are curated at multiple sequence identity thresholds (e.g., 25%, 30%, 40%, etc.) to avoid bias during model training.