Dataset Viewer
Auto-converted to Parquet Duplicate
idx
stringlengths
1
7
smiles
stringlengths
1
111
homolumogap
stringlengths
10
18
SMILES_original
stringlengths
1
111
SMILES_canonical
stringlengths
1
101
structure_key
stringlengths
5
105
structure_status
stringclasses
1 value
wildcard_atom_count
stringclasses
1 value
source_dataset
stringclasses
1 value
source_table
stringclasses
1 value
source_row_first
int64
1
3.38M
observation_count
int64
1
1
split_source
stringclasses
1 value
100015
OC/C(=C(\C(=O)O)/C)/C(=O)O
5.319825777275
OC/C(=C(\C(=O)O)/C)/C(=O)O
C/C(C(=O)O)=C(/CO)C(=O)O
["C/C(C(=O)O)=C(/CO)C(=O)O"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
100,016
1
train
100364
C[C@@H]1CN(F)[C@@H](CN1CC1CC1)C
7.360679656025
C[C@@H]1CN(F)[C@@H](CN1CC1CC1)C
C[C@@H]1CN(CC2CC2)[C@H](C)CN1F
["C[C@@H]1CN(CC2CC2)[C@H](C)CN1F"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
100,365
1
train
100370
C/N=C(\c1ccc(nc1)N/N=C/C(=O)C)/O
4.204158990225
C/N=C(\c1ccc(nc1)N/N=C/C(=O)C)/O
C/N=C(/O)c1ccc(N/N=C/C(C)=O)nc1
["C/N=C(/O)c1ccc(N/N=C/C(C)=O)nc1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
100,371
1
train
100540
C=CC(CC/C=C(/CCC=C(C)C)\C)(C)C
6.566107212565
C=CC(CC/C=C(/CCC=C(C)C)\C)(C)C
C=CC(C)(C)CC/C=C(\C)CCC=C(C)C
["C=CC(C)(C)CC/C=C(\\C)CCC=C(C)C"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
100,541
1
train
100741
O=C1[C@H](C)[C@@H]2N1C(=C(CS2)C(C)C)C
5.60010304329
O=C1[C@H](C)[C@@H]2N1C(=C(CS2)C(C)C)C
CC1=C(C(C)C)CS[C@@H]2[C@@H](C)C(=O)N12
["CC1=C(C(C)C)CS[C@@H]2[C@@H](C)C(=O)N12"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
100,742
1
train
100753
COC(=O)[C@H](Cc1nnn(c1)C(C)C)N
6.492636472930001
COC(=O)[C@H](Cc1nnn(c1)C(C)C)N
COC(=O)[C@@H](N)Cc1cn(C(C)C)nn1
["COC(=O)[C@@H](N)Cc1cn(C(C)C)nn1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
100,754
1
train
100808
O/N=C(/c1cc(ccc1O)C1(C)CC1)\C
4.568791549895
O/N=C(/c1cc(ccc1O)C1(C)CC1)\C
C/C(=N\O)c1cc(C2(C)CC2)ccc1O
["C/C(=N\\O)c1cc(C2(C)CC2)ccc1O"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
100,809
1
train
100993
CC([C@H](C(=O)C(C)C)C[C]1[N]CCN1C)C
5.532074580665
CC([C@H](C(=O)C(C)C)C[C]1[N]CCN1C)C
CC(C)C(=O)[C@H](C[C]1[N]CCN1C)C(C)C
["CC(C)C(=O)[C@H](C[C]1[N]CCN1C)C(C)C"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
100,994
1
train
101057
CC(N1C[C@@H](C[C@H]1C(C)C)n1nccn1)C
5.496699780099999
CC(N1C[C@@H](C[C@H]1C(C)C)n1nccn1)C
CC(C)[C@@H]1C[C@@H](n2nccn2)CN1C(C)C
["CC(C)[C@@H]1C[C@@H](n2nccn2)CN1C(C)C"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
101,058
1
train
101209
OC/C(=C(/C/N=C(/OC(C)(C)C)\O)\C)/F
6.571549489575
OC/C(=C(/C/N=C(/OC(C)(C)C)\O)\C)/F
C/C(C/N=C(\O)OC(C)(C)C)=C(/F)CO
["C/C(C/N=C(\\O)OC(C)(C)C)=C(/F)CO"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
101,210
1
train
101238
C[C@H]1OC=NC[C@@H]1C
7.48313088875
C[C@H]1OC=NC[C@@H]1C
C[C@H]1CN=CO[C@@H]1C
["C[C@H]1CN=CO[C@@H]1C"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
101,239
1
train
101322
C/C=N/OCc1ccc(c(c1)F)F
6.1143982207350005
C/C=N/OCc1ccc(c(c1)F)F
C/C=N/OCc1ccc(F)c(F)c1
["C/C=N/OCc1ccc(F)c(F)c1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
101,323
1
train
101644
CN1CCC[C@@]2([C@H]1c1cc(C)ccc1C2)C
5.281729838204999
CN1CCC[C@@]2([C@H]1c1cc(C)ccc1C2)C
Cc1ccc2c(c1)[C@H]1N(C)CCC[C@@]1(C)C2
["Cc1ccc2c(c1)[C@H]1N(C)CCC[C@@]1(C)C2"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
101,645
1
train
101681
CN1C[C@@H]2[C@H](C1)[C@H]2COS([O])([O])C
6.73753893838
CN1C[C@@H]2[C@H](C1)[C@H]2COS([O])([O])C
CN1C[C@@H]2[C@@H](COS(C)([O])[O])[C@@H]2C1
["CN1C[C@@H]2[C@@H](COS(C)([O])[O])[C@@H]2C1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
101,682
1
train
101686
C[C@H](C(C)(C)C)c1nocn1
7.113056052070001
C[C@H](C(C)(C)C)c1nocn1
C[C@@H](c1ncon1)C(C)(C)C
["C[C@@H](c1ncon1)C(C)(C)C"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
101,687
1
train
101780
CN1[N]N=C([N]1)c1ccccc1F
5.360642854850001
CN1[N]N=C([N]1)c1ccccc1F
CN1[N]N=C(c2ccccc2F)[N]1
["CN1[N]N=C(c2ccccc2F)[N]1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
101,781
1
train
101782
COc1ccccc1C1=N[N]N([N]1)C
4.979683464150001
COc1ccccc1C1=N[N]N([N]1)C
COc1ccccc1C1=N[N]N(C)[N]1
["COc1ccccc1C1=N[N]N(C)[N]1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
101,783
1
train
101832
COc1ncc(c2c1CCO2)[C@@H]1CCCN1C
5.303498946245
COc1ncc(c2c1CCO2)[C@@H]1CCCN1C
COc1ncc([C@@H]2CCCN2C)c2c1CCO2
["COc1ncc([C@@H]2CCCN2C)c2c1CCO2"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
101,833
1
train
101954
O=C(C(C)(C)C)c1cccc2c1CCCO2
4.737502137205
O=C(C(C)(C)C)c1cccc2c1CCCO2
CC(C)(C)C(=O)c1cccc2c1CCCO2
["CC(C)(C)C(=O)c1cccc2c1CCCO2"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
101,955
1
train
101984
OB([C@@H]1CCCN1C(=O)CC[C@H]1CNCC1)O
6.702164137815
OB([C@@H]1CCCN1C(=O)CC[C@H]1CNCC1)O
O=C(CC[C@@H]1CCNC1)N1CCC[C@H]1B(O)O
["O=C(CC[C@@H]1CCNC1)N1CCC[C@H]1B(O)O"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
101,985
1
train
102003
CCOC(=O)/C(=C/Oc1ccc(cc1)Cl)/C
5.202816821560001
CCOC(=O)/C(=C/Oc1ccc(cc1)Cl)/C
CCOC(=O)/C(C)=C/Oc1ccc(Cl)cc1
["CCOC(=O)/C(C)=C/Oc1ccc(Cl)cc1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,004
1
train
102023
C/C(=C\c1cc2c(o1)cccc2)/N
4.17422646667
C/C(=C\c1cc2c(o1)cccc2)/N
C/C(N)=C\c1cc2ccccc2o1
["C/C(N)=C\\c1cc2ccccc2o1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,024
1
train
102030
FCC(/C=C/C(C)C)(C)C
7.26543980835
FCC(/C=C/C(C)C)(C)C
CC(C)/C=C/C(C)(C)CF
["CC(C)/C=C/C(C)(C)CF"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,031
1
train
102041
CC(/C=C/C(NC(=N)O)(C)C)C
6.6559047832300005
CC(/C=C/C(NC(=N)O)(C)C)C
CC(C)/C=C/C(C)(C)NC(=N)O
["CC(C)/C=C/C(C)(C)NC(=N)O"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,042
1
train
102052
CN([C@@H]1CCN(C1)c1ccc2c(c1)CNC2)C
5.085807865845
CN([C@@H]1CCN(C1)c1ccc2c(c1)CNC2)C
CN(C)[C@@H]1CCN(c2ccc3c(c2)CNC3)C1
["CN(C)[C@@H]1CCN(c2ccc3c(c2)CNC3)C1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,053
1
train
102061
CN1CC[C@H](C1)NCC(=O)c1ccccc1
4.334773638465
CN1CC[C@H](C1)NCC(=O)c1ccccc1
CN1CC[C@@H](NCC(=O)c2ccccc2)C1
["CN1CC[C@@H](NCC(=O)c2ccccc2)C1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,062
1
train
102062
CN1C(=O)OC(=O)[C@@H]2[C@@H]1C[C@@H](CC2)C(=O)C
6.168820990835
CN1C(=O)OC(=O)[C@@H]2[C@@H]1C[C@@H](CC2)C(=O)C
CC(=O)[C@@H]1CC[C@@H]2C(=O)OC(=O)N(C)[C@H]2C1
["CC(=O)[C@@H]1CC[C@@H]2C(=O)OC(=O)N(C)[C@H]2C1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,063
1
train
102097
CCO/C(=C\C(=O)c1ccc(cc1)OC)/OC
4.90349158601
CCO/C(=C\C(=O)c1ccc(cc1)OC)/OC
CCO/C(=C\C(=O)c1ccc(OC)cc1)OC
["CCO/C(=C\\C(=O)c1ccc(OC)cc1)OC"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,098
1
train
102143
COC[C@]12O[C@H]([C@H]([C@@H]1O)C(=C(C)C)C2(C)C)C
6.6395779522
COC[C@]12O[C@H]([C@H]([C@@H]1O)C(=C(C)C)C2(C)C)C
COC[C@@]12O[C@@H](C)[C@@H](C(=C(C)C)C1(C)C)[C@@H]2O
["COC[C@@]12O[C@@H](C)[C@@H](C(=C(C)C)C1(C)C)[C@@H]2O"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,144
1
train
102169
CC[C@]12CC(=C)[C@@H]([C@H]2O)CO1
6.54705924303
CC[C@]12CC(=C)[C@@H]([C@H]2O)CO1
C=C1C[C@@]2(CC)OC[C@@H]1[C@H]2O
["C=C1C[C@@]2(CC)OC[C@@H]1[C@H]2O"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,170
1
train
102175
C/C=C/CC(=O)Oc1ccc(cc1)N(=O)=O
4.590560657935001
C/C=C/CC(=O)Oc1ccc(cc1)N(=O)=O
C/C=C/CC(=O)Oc1ccc([N+](=O)[O-])cc1
["C/C=C/CC(=O)Oc1ccc([N+](=O)[O-])cc1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,176
1
train
102195
PC[C@@H]1CCC=C(N1)c1ccccc1S
5.09397128136
PC[C@@H]1CCC=C(N1)c1ccccc1S
PC[C@@H]1CCC=C(c2ccccc2S)N1
["PC[C@@H]1CCC=C(c2ccccc2S)N1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,196
1
train
102334
C#C/C(=C/C=C(/C#C)\C)/C
3.91299717019
C#C/C(=C/C=C(/C#C)\C)/C
C#C/C(C)=C/C=C(\C)C#C
["C#C/C(C)=C/C=C(\\C)C#C"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,335
1
train
102471
CN[C@@H]([C@H](C=C)O)C(=O)C
5.668131505915
CN[C@@H]([C@H](C=C)O)C(=O)C
C=C[C@H](O)[C@H](NC)C(C)=O
["C=C[C@H](O)[C@H](NC)C(C)=O"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,472
1
train
102574
COO/C=N/[C@H]([C@H](OC)C)C(C)C
6.696721860805
COO/C=N/[C@H]([C@H](OC)C)C(C)C
COO/C=N/[C@@H](C(C)C)[C@@H](C)OC
["COO/C=N/[C@@H](C(C)C)[C@@H](C)OC"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,575
1
train
102657
CC[C@@H]1C[C@H]2[C@@H](N1)CC[C@@H](CC2)C
8.06001225181
CC[C@@H]1C[C@H]2[C@@H](N1)CC[C@@H](CC2)C
CC[C@@H]1C[C@@H]2CC[C@@H](C)CC[C@@H]2N1
["CC[C@@H]1C[C@@H]2CC[C@@H](C)CC[C@@H]2N1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,658
1
train
102723
COC(=O)[C@@H]1C[C@H]1C(=O)C(C)C
5.972899018475
COC(=O)[C@@H]1C[C@H]1C(=O)C(C)C
COC(=O)[C@@H]1C[C@H]1C(=O)C(C)C
["COC(=O)[C@@H]1C[C@H]1C(=O)C(C)C"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,724
1
train
102746
N[C@H](C(C(F)(F)F)(C(F)(F)F)O)C(C)C
7.981099235165001
N[C@H](C(C(F)(F)F)(C(F)(F)F)O)C(C)C
CC(C)[C@H](N)C(O)(C(F)(F)F)C(F)(F)F
["CC(C)[C@H](N)C(O)(C(F)(F)F)C(F)(F)F"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,747
1
train
102776
F[C@H]1CNC[C@@H](C1)O[Si](C(C)(C)C)(C)C
7.485852027254999
F[C@H]1CNC[C@@H](C1)O[Si](C(C)(C)C)(C)C
CC(C)(C)[Si](C)(C)O[C@H]1CNC[C@H](F)C1
["CC(C)(C)[Si](C)(C)O[C@H]1CNC[C@H](F)C1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,777
1
train
102803
N[C@H]1CCC[C@H]1[C@@H]1NOC(=N1)S
6.190590098875
N[C@H]1CCC[C@H]1[C@@H]1NOC(=N1)S
N[C@H]1CCC[C@H]1[C@H]1N=C(S)ON1
["N[C@H]1CCC[C@H]1[C@H]1N=C(S)ON1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,804
1
train
102876
CO/C=C\1/CCOc2c1cc(cc2)N(=O)=O
3.872180092615
CO/C=C\1/CCOc2c1cc(cc2)N(=O)=O
CO/C=C1/CCOc2ccc([N+](=O)[O-])cc21
["CO/C=C1/CCOc2ccc([N+](=O)[O-])cc21"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,877
1
train
102934
C[C@@H]1C[C@H]2C[C@@H]([C@@H]1C1=C(C)COCC1)C2(C)C
6.696721860805001
C[C@@H]1C[C@H]2C[C@@H]([C@@H]1C1=C(C)COCC1)C2(C)C
CC1=C([C@@H]2[C@H](C)C[C@H]3C[C@@H]2C3(C)C)CCOC1
["CC1=C([C@@H]2[C@H](C)C[C@H]3C[C@@H]2C3(C)C)CCOC1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,935
1
train
102995
CCOC(=C)/C=C/c1ccc(c(c1)C)C
4.285793145375001
CCOC(=C)/C=C/c1ccc(c(c1)C)C
C=C(/C=C/c1ccc(C)c(C)c1)OCC
["C=C(/C=C/c1ccc(C)c(C)c1)OCC"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
102,996
1
train
103097
CC(=C(/C=C/c1ccsc1C)C)C
4.4762728407250005
CC(=C(/C=C/c1ccsc1C)C)C
CC(C)=C(C)/C=C/c1ccsc1C
["CC(C)=C(C)/C=C/c1ccsc1C"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
103,098
1
train
103222
CC([C@H](C([C@H](N)C)(C)C)C)C
8.11443502191
CC([C@H](C([C@H](N)C)(C)C)C)C
CC(C)[C@@H](C)C(C)(C)[C@@H](C)N
["CC(C)[C@@H](C)C(C)(C)[C@@H](C)N"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
103,223
1
train
103228
OC[C@@H]([C@@H](/C=C/c1ccccc1)O)C
5.08852900435
OC[C@@H]([C@@H](/C=C/c1ccccc1)O)C
C[C@@H](CO)[C@H](O)/C=C/c1ccccc1
["C[C@@H](CO)[C@H](O)/C=C/c1ccccc1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
103,229
1
train
103353
CNC[C@H](C[C@@H](C(=O)O)NC)CC
6.329368162630001
CNC[C@H](C[C@@H](C(=O)O)NC)CC
CC[C@H](CNC)C[C@H](NC)C(=O)O
["CC[C@H](CNC)C[C@H](NC)C(=O)O"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
103,354
1
train
103396
Cc1coc(c1)C(=O)N1C[C@@H]2NC[C@H](C1)C2
5.28445097671
Cc1coc(c1)C(=O)N1C[C@@H]2NC[C@H](C1)C2
Cc1coc(C(=O)N2C[C@H]3CN[C@H](C3)C2)c1
["Cc1coc(C(=O)N2C[C@H]3CN[C@H](C3)C2)c1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
103,397
1
train
103445
CO[CH][N]/N=C(\C1CCCC1)/CC(C)(C)C
5.74160224555
CO[CH][N]/N=C(\C1CCCC1)/CC(C)(C)C
CO[CH][N]/N=C(/CC(C)(C)C)C1CCCC1
["CO[CH][N]/N=C(/CC(C)(C)C)C1CCCC1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
103,446
1
train
103512
CN1CC[C@H](C1)N(CC(=O)N1CCC[C@H]1C)C
6.438213702830001
CN1CC[C@H](C1)N(CC(=O)N1CCC[C@H]1C)C
C[C@@H]1CCCN1C(=O)CN(C)[C@@H]1CCN(C)C1
["C[C@@H]1CCCN1C(=O)CN(C)[C@@H]1CCN(C)C1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
103,513
1
train
103540
CO[C@H](COCc1ccccc1)CCC=C
6.288551085055
CO[C@H](COCc1ccccc1)CCC=C
C=CCC[C@@H](COCc1ccccc1)OC
["C=CCC[C@@H](COCc1ccccc1)OC"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
103,541
1
train
103632
OC[C@H]1O[C@@H](/C(=C\F)/[C@H]1O)C
6.658625921735
OC[C@H]1O[C@@H](/C(=C\F)/[C@H]1O)C
C[C@H]1O[C@H](CO)[C@H](O)/C1=C/F
["C[C@H]1O[C@H](CO)[C@H](O)/C1=C/F"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
103,633
1
train
103679
F[C@H]1C=C(F)c2c(C1=O)nccn2
4.27218745285
F[C@H]1C=C(F)c2c(C1=O)nccn2
O=C1c2nccnc2C(F)=C[C@@H]1F
["O=C1c2nccnc2C(F)=C[C@@H]1F"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
103,680
1
train
103712
O=C1C=C[C]2C(=N1)C(=O)[CH]c1c2nccc1
1.918402646025
O=C1C=C[C]2C(=N1)C(=O)[CH]c1c2nccc1
O=C1C=C[C]2C(=N1)C(=O)[CH]c1cccnc12
["O=C1C=C[C]2C(=N1)C(=O)[CH]c1cccnc12"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
103,713
1
train
103880
CC([C@H]1NCCN(C1)c1ncc(cn1)C)C
4.742944414215
CC([C@H]1NCCN(C1)c1ncc(cn1)C)C
Cc1cnc(N2CCN[C@H](C(C)C)C2)nc1
["Cc1cnc(N2CCN[C@H](C(C)C)C2)nc1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
103,881
1
train
104302
COC(=O)[C@@H]1C=NC(=O)c2c1cccc2F
4.813694015345
COC(=O)[C@@H]1C=NC(=O)c2c1cccc2F
COC(=O)[C@@H]1C=NC(=O)c2c(F)cccc21
["COC(=O)[C@@H]1C=NC(=O)c2c(F)cccc21"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
104,303
1
train
104368
[NH][N][N][N][N][N]N=C=S
3.975583355805
[NH][N][N][N][N][N]N=C=S
[NH][N][N][N][N][N]N=C=S
["[NH][N][N][N][N][N]N=C=S"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
104,369
1
train
104481
CC(/C=C/Cc1ccccc1CC(C)C)C
6.283108808044999
CC(/C=C/Cc1ccccc1CC(C)C)C
CC(C)/C=C/Cc1ccccc1CC(C)C
["CC(C)/C=C/Cc1ccccc1CC(C)C"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
104,482
1
train
104556
CCCC[P]
7.447756088185
CCCC[P]
CCCC[P]
["CCCC[P]"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
104,557
1
train
104829
CC[C@H]1C[C@H]2[C@H](C[C@H]1CC)[C@H](C(=C2C)C)C
6.775634877449999
CC[C@H]1C[C@H]2[C@H](C[C@H]1CC)[C@H](C(=C2C)C)C
CC[C@@H]1C[C@H]2[C@H](C[C@@H]1CC)C(C)=C(C)[C@@H]2C
["CC[C@@H]1C[C@H]2[C@H](C[C@@H]1CC)C(C)=C(C)[C@@H]2C"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
104,830
1
train
104856
CCN1CCC[C@H](C1)C(=O)/C=C/C(C)C
4.31572566893
CCN1CCC[C@H](C1)C(=O)/C=C/C(C)C
CCN1CCC[C@@H](C(=O)/C=C/C(C)C)C1
["CCN1CCC[C@@H](C(=O)/C=C/C(C)C)C1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
104,857
1
train
104874
COCCN1CCC[C@H]1COP(=O)(O)O
6.723933245854999
COCCN1CCC[C@H]1COP(=O)(O)O
COCCN1CCC[C@H]1COP(=O)(O)O
["COCCN1CCC[C@H]1COP(=O)(O)O"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
104,875
1
train
104910
OCc1cc2n[nH]cc2cn1
4.655867982055001
OCc1cc2n[nH]cc2cn1
OCc1cc2n[nH]cc2cn1
["OCc1cc2n[nH]cc2cn1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
104,911
1
train
104915
Clc1ccc2c(c1)n(O)nc2
5.02866395724
Clc1ccc2c(c1)n(O)nc2
On1ncc2ccc(Cl)cc21
["On1ncc2ccc(Cl)cc21"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
104,916
1
train
104977
Nc1cc(OC2CC2)cc(c1)C(=O)O
4.7347809987
Nc1cc(OC2CC2)cc(c1)C(=O)O
Nc1cc(OC2CC2)cc(C(=O)O)c1
["Nc1cc(OC2CC2)cc(C(=O)O)c1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
104,978
1
train
105158
O=Cc1cnc(cc1OC1CC1)C#N
4.566070411390001
O=Cc1cnc(cc1OC1CC1)C#N
N#Cc1cc(OC2CC2)c(C=O)cn1
["N#Cc1cc(OC2CC2)c(C=O)cn1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
105,159
1
train
105213
O=Cc1ccc(cc1C(C)(C)C)OC1CC1
4.802809461325
O=Cc1ccc(cc1C(C)(C)C)OC1CC1
CC(C)(C)c1cc(OC2CC2)ccc1C=O
["CC(C)(C)c1cc(OC2CC2)ccc1C=O"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
105,214
1
train
105228
O=Cc1ccc(c(c1)N(=O)=O)OC1CC1
4.672194813085
O=Cc1ccc(c(c1)N(=O)=O)OC1CC1
O=Cc1ccc(OC2CC2)c([N+](=O)[O-])c1
["O=Cc1ccc(OC2CC2)c([N+](=O)[O-])c1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
105,229
1
train
105242
O=Cc1ccc(c(n1)C(C)C)OC1CC1
4.862674508435001
O=Cc1ccc(c(n1)C(C)C)OC1CC1
CC(C)c1nc(C=O)ccc1OC1CC1
["CC(C)c1nc(C=O)ccc1OC1CC1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
105,243
1
train
105322
O=Cc1c(cccc1N(=O)=O)OC1CC1
4.119803696570001
O=Cc1c(cccc1N(=O)=O)OC1CC1
O=Cc1c(OC2CC2)cccc1[N+](=O)[O-]
["O=Cc1c(OC2CC2)cccc1[N+](=O)[O-]"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
105,323
1
train
105745
NCc1cc(cnc1OC(C)C)OC1CC1
5.18648999053
NCc1cc(cnc1OC(C)C)OC1CC1
CC(C)Oc1ncc(OC2CC2)cc1CN
["CC(C)Oc1ncc(OC2CC2)cc1CN"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
105,746
1
train
105787
NC[C@H]1C=C(C=C[C@@H]1N)OC1CC1
4.97968346415
NC[C@H]1C=C(C=C[C@@H]1N)OC1CC1
NC[C@H]1C=C(OC2CC2)C=C[C@@H]1N
["NC[C@H]1C=C(OC2CC2)C=C[C@@H]1N"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
105,788
1
train
105866
FC1=C(OC2CC2)C=C(N[CH]1)[NH]
4.38103299305
FC1=C(OC2CC2)C=C(N[CH]1)[NH]
[NH]C1=CC(OC2CC2)=C(F)[CH]N1
["[NH]C1=CC(OC2CC2)=C(F)[CH]N1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
105,867
1
train
106003
N#C[C@@H]1[C@H](C=CC=C1C(=O)N)OC1CC1
4.76199238375
N#C[C@@H]1[C@H](C=CC=C1C(=O)N)OC1CC1
N#C[C@H]1C(C(N)=O)=CC=C[C@@H]1OC1CC1
["N#C[C@H]1C(C(N)=O)=CC=C[C@@H]1OC1CC1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
106,004
1
train
106039
OC(=N)c1cc(OC2CC2)cc(c1)C(=O)O
5.004173710695
OC(=N)c1cc(OC2CC2)cc(c1)C(=O)O
N=C(O)c1cc(OC2CC2)cc(C(=O)O)c1
["N=C(O)c1cc(OC2CC2)cc(C(=O)O)c1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
106,040
1
train
106112
N#Cc1ncc(c(c1)OC1CC1)Br
5.221864791095
N#Cc1ncc(c(c1)OC1CC1)Br
N#Cc1cc(OC2CC2)c(Br)cn1
["N#Cc1cc(OC2CC2)c(Br)cn1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
106,113
1
train
106310
O=N(=O)c1c(ccnc1N(=O)=O)OC1CC1
4.857232231425002
O=N(=O)c1c(ccnc1N(=O)=O)OC1CC1
O=[N+]([O-])c1nccc(OC2CC2)c1[N+](=O)[O-]
["O=[N+]([O-])c1nccc(OC2CC2)c1[N+](=O)[O-]"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
106,311
1
train
106543
N#Cc1cc(OC2CC2)cc(c1)C(C)C
5.21914365259
N#Cc1cc(OC2CC2)cc(c1)C(C)C
CC(C)c1cc(C#N)cc(OC2CC2)c1
["CC(C)c1cc(C#N)cc(OC2CC2)c1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
106,544
1
train
106547
N#Cc1cc(ncc1OC1CC1)C(C)(C)C
4.85451109292
N#Cc1cc(ncc1OC1CC1)C(C)(C)C
CC(C)(C)c1cc(C#N)c(OC2CC2)cn1
["CC(C)(C)c1cc(C#N)c(OC2CC2)cn1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
106,548
1
train
106575
N#C[C@@H]1[C@H](N=CC=C1OC1CC1)OC(C)(C)C
5.16472088249
N#C[C@@H]1[C@H](N=CC=C1OC1CC1)OC(C)(C)C
CC(C)(C)O[C@H]1N=CC=C(OC2CC2)[C@@H]1C#N
["CC(C)(C)O[C@H]1N=CC=C(OC2CC2)[C@@H]1C#N"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
106,576
1
train
106578
N#Cc1c(OC2CC2)cncc1N
4.636820012519999
N#Cc1c(OC2CC2)cncc1N
N#Cc1c(N)cncc1OC1CC1
["N#Cc1c(N)cncc1OC1CC1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
106,579
1
train
106722
Clc1cccc(c1OC1CC1)C(F)(F)F
5.90487055585
Clc1cccc(c1OC1CC1)C(F)(F)F
FC(F)(F)c1cccc(Cl)c1OC1CC1
["FC(F)(F)c1cccc(Cl)c1OC1CC1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
106,723
1
train
106863
CSc1ccnc(c1OC1CC1)F
5.249076176145
CSc1ccnc(c1OC1CC1)F
CSc1ccnc(F)c1OC1CC1
["CSc1ccnc(F)c1OC1CC1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
106,864
1
train
106937
CSc1ccc(cc1OC1CC1)F
5.477651810565001
CSc1ccc(cc1OC1CC1)F
CSc1ccc(F)cc1OC1CC1
["CSc1ccc(F)cc1OC1CC1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
106,938
1
train
107045
CSc1ccnc(c1OC1CC1)C#N
4.810972876839999
CSc1ccnc(c1OC1CC1)C#N
CSc1ccnc(C#N)c1OC1CC1
["CSc1ccnc(C#N)c1OC1CC1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
107,046
1
train
107060
CSc1cc(OC2CC2)cc(c1)Cl
5.56200710422
CSc1cc(OC2CC2)cc(c1)Cl
CSc1cc(Cl)cc(OC2CC2)c1
["CSc1cc(Cl)cc(OC2CC2)c1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
107,061
1
train
107091
CSc1cc(OC2CC2)cc(c1)C
5.57289165824
CSc1cc(OC2CC2)cc(c1)C
CSc1cc(C)cc(OC2CC2)c1
["CSc1cc(C)cc(OC2CC2)c1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
107,092
1
train
107171
COc1ncc(cc1OC1CC1)N(=O)=O
4.11980369657
COc1ncc(cc1OC1CC1)N(=O)=O
COc1ncc([N+](=O)[O-])cc1OC1CC1
["COc1ncc([N+](=O)[O-])cc1OC1CC1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
107,172
1
train
107213
COc1cncc(c1C(=O)N)OC1CC1
5.205537960065
COc1cncc(c1C(=O)N)OC1CC1
COc1cncc(OC2CC2)c1C(N)=O
["COc1cncc(OC2CC2)c1C(N)=O"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
107,214
1
train
107509
COc1ccc(c(n1)N(=O)=O)OC1CC1
4.182389882185
COc1ccc(c(n1)N(=O)=O)OC1CC1
COc1ccc(OC2CC2)c([N+](=O)[O-])n1
["COc1ccc(OC2CC2)c([N+](=O)[O-])n1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
107,510
1
train
107510
CNc1ccc(cc1OC1CC1)Cl
5.281729838205
CNc1ccc(cc1OC1CC1)Cl
CNc1ccc(Cl)cc1OC1CC1
["CNc1ccc(Cl)cc1OC1CC1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
107,511
1
train
107753
COC1=C([CH]NC(=C1)[NH])OC1CC1
4.33749477697
COC1=C([CH]NC(=C1)[NH])OC1CC1
COC1=C(OC2CC2)[CH]NC([NH])=C1
["COC1=C(OC2CC2)[CH]NC([NH])=C1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
107,754
1
train
107781
CNc1c(ccnc1C(=O)N)OC1CC1
4.334773638465001
CNc1c(ccnc1C(=O)N)OC1CC1
CNc1c(OC2CC2)ccnc1C(N)=O
["CNc1c(OC2CC2)ccnc1C(N)=O"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
107,782
1
train
107808
COc1ccc(cc1OC1CC1)OC(C)C
5.434113594485
COc1ccc(cc1OC1CC1)OC(C)C
COc1ccc(OC(C)C)cc1OC1CC1
["COc1ccc(OC(C)C)cc1OC1CC1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
107,809
1
train
107852
CNc1cccc(c1C(=O)N(C)C)OC1CC1
4.86539564694
CNc1cccc(c1C(=O)N(C)C)OC1CC1
CNc1cccc(OC2CC2)c1C(=O)N(C)C
["CNc1cccc(OC2CC2)c1C(=O)N(C)C"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
107,853
1
train
107913
COc1ccnc(c1OC1CC1)C=O
4.75655010674
COc1ccnc(c1OC1CC1)C=O
COc1ccnc(C=O)c1OC1CC1
["COc1ccnc(C=O)c1OC1CC1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
107,914
1
train
107932
CNc1cnc(c(c1)OC1CC1)N(C)C
4.69124278262
CNc1cnc(c(c1)OC1CC1)N(C)C
CNc1cnc(N(C)C)c(OC2CC2)c1
["CNc1cnc(N(C)C)c(OC2CC2)c1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
107,933
1
train
108209
CNc1c(ccnc1CC)OC1CC1
5.6817371984400005
CNc1c(ccnc1CC)OC1CC1
CCc1nccc(OC2CC2)c1NC
["CCc1nccc(OC2CC2)c1NC"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
108,210
1
train
108320
CN(c1cncc(c1OC1CC1)C)C
5.161999743985
CN(c1cncc(c1OC1CC1)C)C
Cc1cncc(N(C)C)c1OC1CC1
["Cc1cncc(N(C)C)c1OC1CC1"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
108,321
1
train
108337
CN(c1cncc(c1C)OC1CC1)C
5.281729838205
CN(c1cncc(c1C)OC1CC1)C
Cc1c(OC2CC2)cncc1N(C)C
["Cc1c(OC2CC2)cncc1N(C)C"]
valid
0
pcqm4mv2
pcqm4mv2/data.csv.gz
108,338
1
train
End of preview. Expand in Data Studio

PCQM4Mv2 PubChemQC HOMO-LUMO gap benchmark

Large-scale prediction of quantum-calculated HOMO–LUMO gaps from molecular structures. Includes the official train, validation, test-development, and test-challenge partitions, with a separate archive of training geometries.

Part of the Molecular Datasets collection.

Get started

from datasets import load_dataset
dataset = load_dataset("liuganghuggingface/moldata-pcqm4mv2", "default")

Where the data comes from

Tables

Config Splits Rows Targets
default train, validation, test_challenge, test_dev 3,746,581 homolumogap (regression; eV)

data/ contains rows with readable molecular structures; identical observations are collapsed. observations/ contains all source rows.

Source label columns are text, including values such as <10. SMILES_original retains the input representation; SMILES_canonical provides a canonical representation. structure_key groups molecular components, and structure_status records parsing outcomes.

Files and modalities

Modality Where Content
3D structures structures/pcqm4m-v2-train.sdf.tar.gz (1,560 MB) OGB-LSC DFT-optimized geometries for the training molecules only, unchanged. SDF record i is the molecule with source idx i (formulas agree; bonds perceived from 3D differ from the SMILES graph for about 4% of records). Validation and test molecules have no geometry.

Download an individual file:

from huggingface_hub import hf_hub_download
path = hf_hub_download("liuganghuggingface/moldata-pcqm4mv2", "structures/pcqm4m-v2-train.sdf.tar.gz", repo_type="dataset")

Data quality

  • computed: Labels are computed (for example DFT or molecular dynamics).
  • <0.1% unreadable: 39 rows of 3,746,620 have a SMILES that RDKit cannot read; they are left out of the processed table.
  • 0.3% undefined stereo: 0.3% of molecules have a stereocenter whose configuration is not given, so mirror-image forms cannot be told apart.
  • 7 stereoisomer sets in a sample: Groups of stereoisomers among 20,000 sampled structures. The dataset is too large to compare all their values.
  • R² limit 0.98: Across 111,031 cases of a molecule measured more than once, the spread between repeats is 0.14 times the spread of all values, so even a perfect model reaches an R² of about 0.98. Part of the spread can come from conditions the source does not record.
  • 75 extreme values: 75 of 3,452,118 values lie more than 6 robust standard deviations from the median of their target (on a log scale when values span 1,000-fold or more). They are often typing or unit mistakes, and sometimes real; look at them before training.
  • 3,323,007 labeled molecules: 3,323,007 molecules have a label.
  • source split: train / validation / test challenge / test dev: The source divides the data into train (3,378,573 rows), validation (73,545 rows), test challenge (147,429 rows), test dev (147,034 rows). Use this division to compare with published results.

Dataset notes

  • The benchmark supplies official partitions and withholds test labels.

License

CC-BY-4.0

Citation

Cite the original study and data source:

Downloads last month
374

Collection including liuganghuggingface/moldata-pcqm4mv2

Paper for liuganghuggingface/moldata-pcqm4mv2