Journal of Biomedical Semantics

Table 2 Coverage of term recognition for concepts and relations in experimental data

From: Ranking relations between diseases, drugs and genes for a curation task

PharmGKB data set
		t		tm		tc		tmc
Entity	N	abs	rel	abs	rel	abs	rel	abs	rel
Di	3830	2550	66.58	2872	74.99	2557	66.76	2872	74.99
Dr	4751	3527	74.24	3632	76.45	3668	77.20	3668	77.20
Ge	7522	5838	77.61	5930	78.84	5989	79.62	5994	79.69
TOTAL	16103	11915	73.99	12434	77.22	12214	75.85	12534	77.84
Di-Di	68	22	32.35	24	35.29	22	32.35	24	35.29
Di-Dr	2715	1279	47.11	1484	54.66	1326	48.84	1494	55.03
Di-Ge	5432	3102	57.11	3555	65.45	3181	58.56	3585	66.00
Dr-Dr	181	128	70.72	132	72.93	135	74.59	135	74.59
Dr-Ge	6181	3858	62.42	4016	64.97	4097	66.28	4099	66.32
Ge-Ge	248	141	56.85	142	57.26	145	58.47	145	58.47
TOTAL	14825	8530	57.54	9353	63.09	8906	60.07	9482	63.96
CTD data set
		t		tm		tc		tmc
Entity	N	abs	rel	abs	rel	abs	rel	abs	rel
Di	12639	6939	54.90	9502	75.18	6941	54.92	9502	75.18
Dr	38523	27541	71.49	29531	76.66	30119	78.18	30129	78.21
Ge	39150	28389	72.51	28975	74.01	29169	74.51	29199	74.58
TOTAL	90312	62869	69.61	68008	75.30	66229	73.33	68830	76.21
Di-Ge	6956	4117	59.19	5100	73.32	4163	59.85	5126	73.69
Dr-Di	12154	5335	43.90	8219	67.62	5700	46.90	8356	68.75
Dr-Ge	52746	31015	58.80	33971	64.40	34832	66.04	34883	66.13
TOTAL	71856	40467	56.32	47290	65.81	44695	62.20	48365	67.31

Distribution of identifiable gold standard concepts and relations given the output from our term recognizer and split according to the inclusion of metadata: text only (t), text and MeSH terms (tm), text and chemical substance list (tc), text and all metadata (tmc).

Back to article page

ISSN: 2041-1480

Contact us

General enquiries: journalsubmissions@springernature.com