pydangle top8000: Per-residue backbone geometry for 7,814 quality-filtered protein chains from the MolProbity reference dataset

Per-residue backbone geometry (phi, psi, omega, tau), sidechain torsion angles (chi1-chi4), Ramachandran classifications at four granularities (6/5/4/3-class), DSSP secondary structure, peptide bond type, and chirality for 1,568,561 quality-filtered protein residues from 7,814 chains in 7,607 high-resolution, low-redundancy protein structures. Computed using pydangle-biopython v0.5.1 on "ersatz" full-structure PDB files with NQH flip corrections from Reduce 4.16. Source data from the Richardson Lab Top8000 reference dataset (Williams et al., 2018, doi:10.1002/pro.3330), best-quality, 70% homology. Residue-level filtering by mainchain B-factor <= 30, matching the published Ramachandran contour methodology. See README.md for full methodology and top8000_issues.md for known issues.