Table 1.
Comparison of dihedral and rigid-group parameterizations of conformational change through crystallographic cross-validation.
| Model | R | Rfree |
|---|---|---|
| Arginine kinase, substrate-free (apo-AK), full refined, 1.7 Å resolution (Niu et al., 2011) | 0.189 | 0.241 |
| apo-AK, without residues disordered in transition state analog structure (TSA), waters & TLS-based anisotropic B-factors | 0.262 | 0.287 |
| TSA structure (Yousef et al., 2000) superposed as a rigid subunit | 0.534 | 0.538 |
| TSA structure (Yousef et al., 2000) superposed with DynDom (Hayward and Lee, 2002), default parameters yielding 2 quasi-rigid domains | 0.505 | 0.510 |
| TSA-superposed, manually varying DynDom parameters to yield 3 or 4 dynamic domains | 0.510±0.001 | 0.520±0.003 |
| TSA-superposed by ESCET distance difference matrix analysis (Schneider, 2002), yielding 5 rigid groups | 0.457 | 0.458 |
| TSA-superposed, rigid clusters combined from DynDom and ESCET and manually consolidated into consensus 5 groups | 0.466 | 0.458 |
| TSA-superposed by parsimonius dihedral rotation with tight restraint, λ = 5.0 | 0.453 | 0.450 |
| TSA-superposed by parsimonius dihedral rotation, looser restraint, λ = 2.0 | 0.431 | 0.435 |