Development of a Machine Learning Method to Predict Membrane Protein-Ligand Binding Residues Using Basic Sequence Information

Table 2

Per protein prediction accuracy.

Pdb ID	Number of residues	Sensitivity	Specificity	Accuracy
1ar1_b	252	50	83.2	79.44
1be3_e	196	80	89.6	89.06
1ds8_h	246	40	91.6	90.50
1ds8_l	281	54.8	65.2	62.45
1ehk_a	544	43.5	71	66.30
1ehk_b	166	30	77	74.07
1eys_m	318	52.6	75.5	67.20
1f88_a	338	25.5	55.9	50.90
1fx8_a	254	29.5	68	61.2
1iwo_a	994	10	82.3	80.81
1j4n_a	249	50	68.2	67.34
1j95_a	98	60	62.9	62.77
1kb9_c	385	66.7	79.4	74.80
1kb9_h	93	20	73.8	70.78
1kb9_i	55	100	38.6	47.06
1kf6_c	130	100	19.8	23.02
1kf6_d	119	36.8	62.5	58.26
1kpl_a	430	23.8	72.8	70.43
1kqf_b	289	47.8	88.4	78.60
1lgh_a	56	79.4	16.7	57.69
1lnq_a	301	0	79.8	78.45
1m0k_a	222	59.5	68.3	65.13
1m56_a	547	50.4	69.1	65.19
1m56_b	260	33.3	79.9	77.73
1nek_c	129	92.3	27.9	48
1nek_d	113	92.6	3.7	25.69
1okc_a	292	27.5	88.5	71.53
1oy9_a	1006	100	84	84.03
1p49_a	549	44.4	73.1	70.11
1p7b_a	258	60	76.7	76.38
1ppj_g	75	14.3	64.9	54.93
1pv7_a	417	37.5	65.7	65.13
1q16_c	224	57.1	42.9	38.89
1q90_a	292	35.1	89.6	82.64
1q90_b	212	50	78.3	66.35
1qle_c	273	35.5	65.5	62.08
1rc2_b	231	40	63.2	61.67
1v54_d	144	84.2	42.1	47.86
1v54_g	84	60	20	40
1v54_j	58	30	81.8	72.22
1vf5_b	138	42.4	83.2	73.13
1vf5_d	168	61.1	82.9	80.49
Average		50.18	66.04	65.25