AlphaFlex Ensembles of the Canonical Human Proteome Representing Intrinsically Disordered Protein Regions

Full-length atomistic ensembles generated to-date using the AlphaFlex workflow for the entire canonical human proteome with intrinsically disordered protein regions. The current template for AlphaFlex ensembles are based on the AlphaFold 9606 Human v4 database from July 2024. The "AlphaFlex_finished_ids_Nov2025.csv" file contains all of the UniProt IDs in the current version of the dataset along with their category numbers and method of generation (AFX-IDPCG/AFX-IDPForge). An example ensemble file for a Category 3 AFX-IDPCG ensemble "hSHANK3_Q9BYB0_idpcg_n100.pdb" that was analyzed in the manuscript has been provided. The "AlphaFlex_database_Jul2024.json" file contains the database with IDR boundaries and mean PAE interaction scores between folded (F) and disordered (D) domains. Where folded domain interactions are only logged when the mean PAE between FX-FY is less than 15 (where X and Y are domain numbers). For the best download speed performance for large datasets, please consider using the FireFox web-browser.