LoCo: a novel main chain scoring function for protein structure prediction based on local coordinates

BMC Bioinformatics

Table 4 Function comparison: decoy discrimination

	R_B1	R_B10	RMSD_decoy	Z_decoy	CC_decoy	FE_decoy (%)	log(P_B1)	log(P_B10)
LoCo	154.8	5.6	3.51	0.938	0.505	31.4	-0.864	-1.640
DFMAC	108.9	13.8	3.64	1.024	0.533	31.6	-0.825	-1.586
RF_CB_SRS_OD	172.8	52.5	4.11	0.914	0.457	28.4	-0.761	-1.524
ProSa 2003	118.2	24.8	3.82	0.931	0.493	32.3	-0.755	-1.650
Four-body	124.8	36.8	5.01	0.539	0.328	19.6	-0.488	-1.317
General-four-body	186.3	52.2	4.97	0.436	0.312	17.1	-0.482	-1.241
Short-range	192.4	36.6	5.27	0.377	0.267	14.9	-0.536	-1.249
BFKV	139.4	32.3	3.98	0.673	0.398	25.3	-0.671	-1.412
BT	175.4	39.0	4.32	0.636	0.391	24.3	-0.558	-1.406
GKS	164.6	71.3	5.57	0.187	0.231	13.0	-0.400	-1.166
HLPL	161.3	31.7	3.86	0.725	0.396	27.3	-0.650	-1.434
MJ1	183.1	34.9	4.54	0.596	0.336	22.5	-0.594	-1.302
MJ2h	149.5	28.4	3.93	0.695	0.407	25.1	-0.640	-1.361
MJ3	172.4	25.3	4.92	0.270	0.241	15.8	-0.476	-1.241
MJ3h	152.2	26.5	4.00	0.641	0.416	25.7	-0.615	-1.417
MJPL	129.5	29.0	4.15	0.693	0.375	25.6	-0.650	-1.363
MS	191.1	67.1	5.09	0.217	0.236	13.4	-0.496	-1.155
MSBM	140.1	65.4	5.77	-0.001	0.165	7.6	-0.359	-1.043
Qa	176.9	32.0	5.02	0.357	0.286	17.2	-0.480	-1.323
Qm	189.6	53.8	5.02	0.356	0.268	16.9	-0.479	-1.224
Qp	163.9	31.4	3.81	0.721	0.377	25.0	-0.622	-1.381
RO	231.3	44.3	5.83	0.197	0.240	15.5	-0.396	-1.194
SKJG	146.4	41.1	4.45	0.504	0.362	20.4	-0.586	-1.309
SKOa	90.7	37.1	4.46	0.469	0.352	20.8	-0.593	-1.328
SKOb	144.3	27.9	4.34	0.573	0.363	21.8	-0.587	-1.395
TD	156.4	30.6	4.17	0.724	0.416	26.3	-0.604	-1.418
Tel	141.3	64.4	4.26	0.540	0.363	22.0	-0.569	-1.402
TEs	163.3	62.1	4.74	0.523	0.338	18.5	-0.522	-1.237
TS	125.5	28.7	4.04	0.697	0.381	26.3	-0.656	-1.358
VD	149.8	48.4	5.22	0.504	0.292	17.3	-0.435	-1.284

Comparison of decoy discrimination performance among all tested functions is shown. All reported measures are averages over the 77 decoy sets in the final testing group. Lower scores are better for R_B1, R_B10, RMSD_decoy, log(P_B1) and log(P_B10). Higher scores are better for Z_decoy, CC_decoy and FE_decoy. LoCo outperforms all other functions in R_B10, RMSD_decoy, and log(P_B1). It is slightly higher than ProSa 2003 in log(P_B10) and slightly lower than ProSa 2003 in FE_decoy. Z_decoy, CC_decoy and FE_decoy for LoCo are all slightly lower than for DFMAC. LoCo outperforms the remaining 27 functions in every measure except R_B1. All metrics are defined in Performance measures at the end of Methods.

ISSN: 1471-2105