Skip to content

MolMini small, chembl_test_linux

Failure breakdown

Attempted: 3000 | exact: 2570 (0.857)

failure mode n share of all share of failures what it points at
unparseable output 7 0.002 0.016 decoder emitted a string RDKit cannot read
stereochemistry only 210 0.070 0.488 skeleton correct; wedge/hash reading wrong
right formula, wrong bonds 7 0.002 0.016 atoms counted correctly, connectivity misread
other structural error 206 0.069 0.479 a genuinely different molecule

Accuracy by molecule size

heavy atoms n exact valid
<=10 9 1.000 1.000
<=15 84 0.833 1.000
<=20 324 0.907 0.997
<=25 741 0.901 0.997
<=30 741 0.879 0.997
<=35 576 0.842 0.997
<=40 327 0.801 1.000
>40 198 0.662 1.000

Accuracy by rendering style

style n exact valid
clean 1000 0.917 0.999
degraded 1000 0.748 0.995
varied 1000 0.905 0.999

Sequence length

  • Reference SMILES length, all attempted: median 47 tokens
  • Reference SMILES length, correct predictions: median 46 tokens

Longest failures

reference predicted
O=S(=O)(O)c1ccc(-c2ccc(S(=O)(=O)Nc3cc(S(=O)(=O)O)cc4cc(S(=O)(=O)O)cc(S O=S(=O)(O)c1ccc(-c2ccc(S(=O)(=O)O)cc2)cc1
CC(=O)OC[C@H]1O[C@@H](O[C@H]2[C@H](OC(C)=O)[C@@H](OC(C)=O)[C@H](S(N)(= CC(=O)OC[C@H]1O[C@@H](OC(C)=O)[C@H](OC(C)=O)[C@@H](OC(C)=O)[C@@H]1O[C@
CC(=O)OC[C@H]1O[C@@H](O[C@H]2[C@H](OC(C)=O)[C@@H](OC(C)=O)[C@H](S(N)(= CC(=O)OC[C@H]1O[C@@H](S(N)(=O)=O)[C@H](OC(C)=O)[C@@H](OC(C)=O)[C@@H]1O
CC(=O)OC[C@H]1O[C@@H](O[C@H]2[C@H](OC(C)=O)[C@@H](OC(C)=O)[C@H](S(N)(= CC(=O)O[C@H]1CC[C@@]2(C)C(=CC[C@H]3[C@@H]4CC[C@H](C(C)C)[C@@]4(C)CC[C@
Cc1c(C)c2c(c(C)c1OC(=O)CCC(=O)NC1CC(C)(C)N([O])C1(C)C)CCC(C)(C(=O)NC1C Cc1c(C)c2c(c(C)c1OC(=O)CCC(=O)NC1CC(C)(C)N([O])C1(C)C)CCC(C)(C(=O)NC1C