Table 1.

Leave-one-out prediction performances of the Echo algorithm compared with the other alternatives

	Cutoff	Sn	Sp	Ac	MCC	Cutoff	Sn	Sp	Ac	MCC
	Echo					GPS 2.1
MAPK3	2.66	0.6593	0.9598	0.9558	0.3321	High	0.6437	0.9537	0.9492	0.3104
	2.22	0.9451	0.9214	0.9218	0.3485	Medium	0.9310	0.8939	0.8944	0.3027
	2.2	0.9560	0.9198	0.9203	0.3492	Low	0.9425	0.8451	0.8464	0.2503
MAPK8	3	0.4848	0.9781	0.9658	0.4027	High	0.3056	0.9648	0.9497	0.2032
	2.08	0.9697	0.9157	0.9170	0.4490	Medium	0.9444	0.9128	0.9135	0.4150
	1.82	1.0000	0.9071	0.9094	0.4438	Low	0.9444	0.8574	0.8594	0.3264
CDK5	2.3	1.0000	0.9672	0.9678	0.5931	High	0.6316	0.9627	0.9568	0.3684
						Medium	1.0000	0.9206	0.9220	0.4141
						Low	1.0000	0.8651	0.8675	0.3205
EGFR	1.76	0.6909	0.9471	0.9159	0.6191	High	0.6393	0.9410	0.9056	0.5610
	1.53	0.7818	0.9169	0.9004	0.6108	Medium	0.7377	0.8734	0.8574	0.4934
	1.44	0.8727	0.8992	0.8960	0.6374	Low	0.7705	0.8013	0.7977	0.4268
SYK	1.5	0.6735	0.9440	0.9113	0.5970	High	0.5667	0.9386	0.9086	0.4543
	1.3	0.9184	0.9188	0.9187	0.7064	Medium	0.8824	0.8910	0.8900	0.6289
	1.24	0.9184	0.8936	0.8966	0.6559	Low	0.9020	0.8256	0.8349	0.5408
Met	0.94	1.0000	0.9505	0.9606	0.8930	High	0.9615	0.9223	0.9302	0.8126
						Medium	1.0000	0.8252	0.8605	0.6983
						Low	1.0000	0.7961	0.8372	0.6636
PKA_alpha	1.9030	0.7188	0.9804	0.9767	0.4870	High	0.6541	0.9769	0.9714	0.4534
	1.8350	0.8125	0.9603	0.9582	0.4154	Medium	0.8054	0.9415	0.9391	0.3775
	1.8100	0.9375	0.9388	0.9388	0.3957	Low	0.8946	0.9032	0.9030	0.3310
MAPK1	2.3426	0.8333	0.9592	0.9573	0.4264	High	0.7304	0.9533	0.9500	0.3561
	2.0110	0.9250	0.9250	0.9250	0.3608	Medium	0.9130	0.8990	0.8992	0.3090
	2.0096	0.9333	0.9249	0.9250	0.3639	Low	0.9304	0.8470	0.8482	0.2526
Abl	1.6458	0.6250	0.9447	0.9170	0.5234	High	0.4375	0.9170	0.8745	0.3164
	1.6300	0.6458	0.9328	0.9079	0.5058	Medium	0.5208	0.8644	0.8339	0.2915
	1.6000	0.6667	0.9308	0.9079	0.5155	Low	0.5833	0.7854	0.7675	0.2429
PKG	2.2300	0.5909	0.9877	0.9806	0.5141	High	0.5238	0.9866	0.9783	0.4554
	2.2000	0.7273	0.9770	0.9726	0.5026	Medium	0.6190	0.9641	0.9579	0.3673
	2.1000	0.7273	0.9659	0.9617	0.4352	Low	0.6905	0.9360	0.9316	0.3145
Aurror_A	1.3600	0.5517	0.9836	0.9719	0.5028	High	0.3214	0.9746	0.9573	0.2657
	1.3000	0.6897	0.9326	0.9260	0.3641	Medium	0.5714	0.9317	0.9221	0.2956
	1.2130	0.7586	0.9123	0.9082	0.3546	Low	0.6071	0.8888	0.8813	0.2417
ATM	1.7281	0.9123	0.9641	0.9633	0.4885	High	0.8246	0.9628	0.9607	0.4389
	1.7157	1.0000	0.9628	0.9633	0.5238	Medium	0.9649	0.9474	0.9477	0.4413
	1.7000	1.0000	0.9617	0.9623	0.5184	Low	0.9825	0.9443	0.9448	0.4381
	Echo					SUMOsp
SUMO	3.1838	0.9015	0.9965	0.9923	0.9084	Medium	0.8817	0.9260	0.9243	0.5060
	2.038	0.9284	0.8879	0.8897	0.4731	Low	0.9247	0.8545	0.8572	0.3933
	3.237	0.8925	0.9975	0.9928	0.9138	High	0.8065	0.9670	0.9609	0.6128
	Echo					GPS 3.0
Nitrated Y	1.9193	0.3125	0.9260	0.8833	0.2114	High	0.2889	0.9002	0.8257	0.1884
	1.7	0.5521	0.8684	0.8464	0.2912	Medium	0.4053	0.8502	0.7960	0.2171
						Low	0.5009	0.8018	0.7651	0.2335
S nitro	2.65	0.3202	0.9921	0.91289	0.48727	High	0.2520	0.9117	0.8040	0.1897
	2	0.48276	0.94273	0.8885	0.44309	Medium	0.3532	0.8672	0.7833	0.2175
	1.69	0.54187	0.83608	0.80139	0.30003	Low	0.5357	0.8014	0.7580	0.2864

The cutoff values of Echo are tuned to match the similar specificity levels of the alternative algorithms, and the four performance measurements are calculated for the Echo algorithm. The performances of the alternative algorithms are collected from the respective publications. Rows of “SUMO” are for SUMOylation, rows of “Nitrated Y” are for Nitrated tyrosine, and rows of “S nitro” are for S nitrosylation