Table 3.

Potential analogues of methyleugenol, similarity scores and availability of human or rodent PBK models.

Chemical	Tanimoto similarity score (OECD QSAR Toolbox)	Average similarity score (KNIME workflow)	Reference for human PBK model	Reference for rodent PBK model	Validity / reproducibility of PBK model
Eugenol	0.82	(excluded as full model unavailable)	[27]	–	Irreproducible - full model not reported.
Elemicin	0.80	0.78	[28] PBK model developed using read across approach	[28] PBK model developed using read across approach	Not validated with in vivo data
Estragole	0.70	0.66	[14], [15], [20]	[29], [30], [31]	Valid and reproducible
Safrole	0.64	0.69	[16]	[16]	Valid and reproducible
Myristicin	0.66	0.58	[32] PBK model developed using read across approach	[32] PBK model developed using read across approach	Not validated with in vivo data
Apiol	< 0.6	0.49	[33]	[33]	Not validated with in vivo data
Isosafrole	< 0.6		Not available	Not available	–
Anethole	< 0.6		Not available	Not available	–
Ally benzene	< 0.6		Not available	Not available	–
Methyleugenol*	1.00	1.00	[19]	[34]	Valid and reproducible*

^*Reference to methyleugenol is for completeness; this model was used to validate the approach, demonstrate the proof-of principle after characterisation and parameterisation of the model derived from analogues.