FuzzyBench-NOE: A Curated NMR Distance Restraint Dataset for Fuzzy IDP–Target Protein Complexes

Overview This dataset provides the first comprehensive collection of NMR distance restraint files for fuzzy protein complexes, assembled to support benchmarking of AI structure prediction methods on intrinsically disordered protein (IDP)–target interactions. Fuzzy complexes are biomolecular assemblies where the IDP retains conformational disorder upon binding. Crystal structures capture only the most ordered snapshot of these ensembles — making standard structure prediction benchmarks misleading. NOE distance restraints, measured in solution by NMR spectroscopy, provide a model-independent experimental ground truth that reflects the actual solution ensemble. Dataset Contents restraints.zip ├── systems_metadata.csv # PDB ID, BMRB ID, filename mapping ├── NOE_restraints_files.zip/ # 75 NMR-STAR restraint files ├── merged_15072_2jmx.str ├── merged_15357_2khs.str └── ... (75 files total) ├── restraint_violation_analysis.py # violation scoring script Curation Criteria Systems were selected to meet all of the following criteria: 1. FuzDB inclusion — the complex is catalogued in FuzDB (v4.0), which requires experimental evidence of bound-state disorder. Disorder-to-order transitions (where the IDP fully folds upon binding) are excluded by FuzDB's curation criteria. 2. PDB structure available — a deposited structure exists in the RCSB Protein Data Bank, either crystal or NMR. 3. BMRB NOE restraints available — a corresponding BMRB entry exists with distance restraint data in NMR-STAR format. Entries were manually verified to contain usable inter-molecular or intra-molecular NOE distance restraints. 4. IDP binds primarily via α-helical motifs — required for compatibility with the Hadži helix–coil thermodynamic framework used in the associated benchmarking analysis. Systems that met criteria 1–2 but lacked BMRB restraint data (29 systems) are included in the broader benchmark dataset (`systems_metadata.csv`) but are not represented in the restraint files. systems_metadata.csv | Column | Description | |--------|-------------| | `pdb_id` | 4-character RCSB PDB identifier | | `bmrb_id` | BMRB entry number (restraint source) | | `str_filename` | Filename of the NMR-STAR restraint file | | `has_noe` | Whether NOE restraint file is included (True/False) | | `fuzdb_entry` | FuzDB entry URL | | `pdb_url` | RCSB PDB entry URL | | `bmrb_url` | BMRB entry URL (where applicable) | Chain and Residue Mapping The script automatically detects chain IDs and residue numbering offsets between the NMR-STAR file and the predicted structure using residue-name matching. Of the 75 systems in this dataset, 74 were handled fully automatically. One system (2MLZ) required a manual chain assignment override due to non-standard numbering in the deposited restraint file; the override is documented in the script's `STRUCTURE_REMAP` configuration block. License This dataset is released under the **Creative Commons Attribution 4.0 International (CC BY 4.0)** license. You are free to share and adapt the material for any purpose, provided appropriate credit is given. Contact Taseef Rahman — University of South Florida Bellini College of AI, Computing, and Cybersecurity [e-mail]