G-Aligner: a graph-based feature alignment method for untargeted LC–MS-based metabolomics

BMC Bioinformatics

Table 6 Evaluation results on software self-extracted features on the MTBLS562 dataset

	TP	FP	TN	FN	P	R	F	F_ACC	A_ACC
MZmine2 RANSAC	5751	106	2398	25	0.982	0.996	0.989	0.984	0.768
Local bipartite	5719	58	2448	55	0.990	0.990	0.990	0.986	0.845
G-Aligner VLSNS_MSR	5744	67	2435	34	0.988	0.994	0.991	0.988	0.850
G-Aligner VLSNS_MSG	5758	66	2434	22	0.989	0.996	0.992	0.989	0.845
OpenMS QT	3916	139	4182	43	0.966	0.989	0.977	0.978	0.715
Local bipartite	3933	32	4277	38	0.992	0.990	0.991	0.992	0.894
G-Aligner VLSNS_MSR	3961	34	4275	10	0.991	0.997	0.994	0.995	0.899
G-Aligner VLSNS_MSG	3966	35	4275	4	0.991	0.999	0.995	0.995	0.899
XCMS Group	5786	386	1671	437	0.937	0.930	0.934	0.901	0.676
XCMS OBI-Warp	5756	437	1671	416	0.929	0.933	0.931	0.897	0.652
Local bipartite	6425	37	1807	11	0.994	0.998	0.996	0.994	0.923
G-Aligner VLSNS_MSR	6425	38	1807	10	0.994	0.998	0.996	0.994	0.918
G-Aligner VLSNS_MSG	6431	37	1807	5	0.994	0.999	0.997	0.995	0.923

The results with the highest performance in the comparison are indicated in bold

ISSN: 1471-2105