Table 1.

Short presentation of the computational tools that were used in the present study. QSAR – Quantitative Structure-Activity Relationship

Tool	Method	Output	Accuracy of predictions, %	References
SwissADME	expert-rules based 2D QSAR	Drug likeness, pharmacokinetic profile specifying Yes or NO for every investigated biological action	72-94	[14]
admetSAR2.0	2D QSAR	Pharmacokinetic profiles, organ (eye, heart, liver) and genomic toxicity specifying the probability of the presence or absence of a biological action.	72-77	[18, 19]
Pred-hERG	2D QSAR	Ability of a chemical compound to inhibit the human ether-à-go-go related gene (hERG)K+ channels using both a binary and a multiclass model.	70-89	[21, 30, 31]
Pred-Skin	2D-QSAR	Skin sensitization potential based on multiple QSAR models: prediction by binary model using human data, binary and multiclass predictions of murine skin sensitization potential based on animal data, and binary predictions based on non-animal data, i.e Direct Peptide Reactivity Assay (DPRA), KeratinoSens, and the human Cell Line Activation Test (h-CLAT)]	70-84	[20, 21, 31]
Endocrine Disruptome	Molecular docking	Probability of binding to nuclear receptors.	70-90	[22]
Toxtree	expert-rules based	Carcinogenic and mutagenic potential expressed by Yes or No.	70	[23]
CarcinoPred-EL	2D QSAR	Carcinogenic potential expressed by Yes or No.	70	[24]