Skip to main content

Table 1 The Ames and Tox21 datasets

From: Analysis of the benefits of imputation models over traditional QSAR models for toxicity prediction

Ames

Tox21

Assay name

Number labels (proportion)

Proportion actives

Assay name

Number labels (proportion)

Proportion actives

TA100

4627 (0.75)

0.26

NR-AhR

6810 (0.84)

0.12

TA100_S9

4350 (0.71)

0.32

NR-AR

7460 (0.92)

0.03

TA102

880 (0.14)

0.17

NR-AR-LBD

6991 (0.86)

0.03

TA102_S9

763 (0.12)

0.21

NR-Aromatase

6009 (0.74)

0.05

TA1535

2489 (0.40)

0.11

NR-ER

6367 (0.79)

0.11

TA1535_S9

2347 (0.38)

0.12

NR-ER-LBD

7199 (0.89)

0.04

TA1537

2081 (0.34)

0.12

NR-PPAR-gamma

6752 (0.83)

0.03

TA1537_S9

1998 (0.32)

0.12

SR-ARE

6121 (0.76)

0.16

TA97

1049 (0.17)

0.14

SR-ATAD5

7326 (0.91)

0.04

TA97_S9

1010 (0.16)

0.17

SR-HSE

6794 (0.84)

0.05

TA98

4345 (0.70)

0.24

SR-MMP

6074 (0.75)

0.15

TA98_S9

4055 (0.66)

0.29

SR-p53

7049 (0.87)

0.06

overall

6168 (0.41)

0.21

overall

8090 (0.83)

0.07

  1. Each row contains the total number of molecules with an experimental label for the assay as well the proportion of active labels for the given assay. The last row (‘overall’) reports the number of unique compounds across all the assays (after the data processing steps described in ‘Data processing’). The proportion of labels in the last row indicates the completeness of the data table across all assays. The proportion of actives in the last row reports the proportion of actives among all available labels.