Comparison of seven in silico tools for evaluating of daphnia and fish acute toxicity: case study on Chinese Priority Controlled Chemicals and new chemicals

Table 8 Tool performance to 96 h-LC₅₀ of fish for chemicals within each applicability domains

Chemicals	Methods	Measures of predictive accuracy	ECOSAR	T.E.S.T.	Danish Q. D.	VEGA	Read Across	Trend Analysis	KATE
37 PCCs	General	Number of inside AD	29	22	19	29	31	30	22
	General	Number of outside AD and missing prediction	8	15	18	8	6	6	15
	Qualitative	Number of correct	16	9	11	16	11	8	8
		Number of incorrect	13	13	8	13	20	23	14
		Accuracy inside AD (%)	55	41	58	55	35	26	36
		R²_AD (toxicity class)	0.66	0.41	0.58	0.57	0.07	0.06	0.35
	Quantitative	Accuracy within a factor of 10 (%)	83	82	74	90	55	42	86
		Accuracy within a factor of 100 (%)	100	95	79	97	81	55	95
		Accuracy within a factor of 1000 (%)	100	100	84	100	94	84	100
		RMSE (log10 LC₅₀)	0.71	0.87	1.83	0.75	1.47	2.09	0.80
		R²_AD (log10 LC₅₀)	0.68	0.52	0.57	0.68	0.14	0.00	0.50
86 NCs	Qualitative	Number of inside AD	58	61	22	50	/	/	21
		Number of outside AD and missing prediction	28	25	64	36	/	/	67
		Number of correct	32	25	13	18	/	/	7
		Number of incorrect	26	36	9	32	/	/	12
		Accuracy inside AD (%)	55	41	59	36	/	/	37
		R²_AD (toxicity class)	0.37	0.10	0.42	0.03	/	/	0.37

ISSN: 1471-2105