Table 1.

Percentage of cases for which at least one of experimental condition component combinations were found within the top of a sorted list of predicted conditions (retrospective test set). LRM—likelihood ranking model, NM—null model.

Mode	TOP 1		TOP 3		TOP 5		TOP 10
	LRM	NM	LRM	NM	LRM	NM	LRM	NM
catalyst	82	66	93	75	96	85	98	91
pressure	78	59	100	100	100	100	100	100
additives	91	87	99	97	99	100	99	100
temperature	93	85	100	100	100	100	100	100
catalyst, pressure	67	44	85	66	90	71	95	81
additives, pressure	72	52	93	87	97	92	99	98
catalyst, additives	76	58	89	67	92	76	96	84
catalyst, pressure	67	44	85	66	90	71	95	81
catalyst, temperature	78	60	90	70	93	76	96	86
temperature, pressure	75	56	93	85	98	97	100	100
catalyst, pressure, additives	63	39	80	58	86	64	91	73
catalyst, temperature, additives	73	53	86	61	89	69	93	80
catalyst, temperature, pressure	65	42	81	60	86	65	91	75
temperature, pressure, additives	69	50	88	74	93	84	97	94
catalyst, additives, pressure, temperature	61	38	77	53	82	59	88	67

LRM—Likelihood ranking model, NM—null model.