new PDB depositions vs. their blind AlphaFold predictions — a running test of "is folding solved?"

13CM_G

Prothrombin · P00734 · RCSB 13CM · AF-P00734-F1 (v6)

0.8682
TM-score
0.8574
lDDT
1.419
Cα-RMSD Å
81.76
mean pLDDT
0.0541
FRAUD score
100.0%
novelty (max id)

Per-domain breakdown

SourceDomainRangeResiduesTMRMSD Åmean Cα ΔName
PAEPAE:1-6221-622 620.8511 1.4191.092

All metrics

Global fold agreement

TM-score (norm. experiment)0.8682
TM-score (norm. model)0.0966
TM-score (norm. shorter)0.8682
TM-score (norm. longer)0.0966
Cα-RMSD (Å)1.419
backbone-RMSD (Å)1.368
all-atom-RMSD (Å)1.828
core-RMSD (Å)0.677
core fraction0.839
GDT_TS88.71
GDT_HA66.94
MaxSub0.8708
structural overlap (3.5 Å)0.9677

Local, superposition-free

lDDT0.8574
contact-map Jaccard0.7619
contact precision0.8205
contact recall0.9143
distance-matrix mean Δ (Å)0.65
CAD-score (approx)0.8143

Backbone & secondary structure

SS agreement Q3 (%)83.87
mean Δφ (°)23.3
mean Δψ (°)20.4
torsion within 30° (frac)0.6774
Rg experiment (Å)11.47
Rg model (Å)11.36
ΔRg (Å)0.12

Confidence calibration

mean pLDDT81.76
pLDDT↔lDDT Pearson0.7564
pLDDT↔lDDT Spearman0.4807
PAE↔observed Pearson0.5545
PAE overconfident frac0.0
mean PAE (Å)5.552
mean observed error (Å)0.639

Context & headline

coverage of model0.0997
coverage of experiment1.0
seq identity aligned (%)83.87
confidently-wrong residue frac0.0
FRAUD score0.0541

Deposited 2026-04-29 · released 2026-07-08 · EM · 3.2 Å · closest pre-cutoff chain: 4NZQ_1