Live Stats, next update: Wed 02 Sep
Human PDBs Analysed
Confidently Wrong
Novel + Confidently Wrong
DB size
Visitors
Full statistics →
New PDB Depositions vs. Their Blind AlphaFold Predictions — A Running Test of “Is Folding Solved?”

Worst offenders — all time

Grouped by protein (AlphaFold has one model per sequence, so every deposition of a protein shares it). Worst first — confidently wrong, then by FRAUD score. Tap a protein to expand its individual structures.

82,901
total analysed
1,9482.3%
confidently wrong
1,9182.3%
novel sequences
0.954
median TM-score
Protein ×N TM pLDDT FRAUD