Benchmark study of feature selection strategies for multi-omics data

BMC Bioinformatics

Table 3 Summary of the datasets used for the benchmark experiment

Dataset	Cancer	Clin	cnv	mirna	mutation	rna	f	n	m	r_m
BLCA	Bladder urothelial	5	57,964	825	18,577	23,081	100,455	382	186	0.49
BRCA	Breast invasive C	8	57,964	835	17,975	22,694	99,479	735	255	0.35
COAD	Colon AC	7	57,964	802	18,538	22,210	99,524	191	106	0.55
ESCA	Esophageal C	6	57,964	763	12,628	25,494	96,858	106	83	0.78
HNSC	Head–neck squamous CC	11	57,964	793	17,248	21,520	97,539	443	307	0.69
LGG	Low grade glioma	10	57,964	645	9235	22,297	90,154	419	195	0.47
LIHC	Liver hepatocellular C	11	57,964	776	11,821	20,994	91,569	159	44	0.28
LUAD	Lung AC	9	57,964	799	18,388	23,681	100,844	426	212	0.50
LUSC	Lung squamous CC	9	57,964	895	18,500	23,524	100,895	418	346	0.83
PAAD	Pancreatic AC	10	57,964	612	12,392	22,348	93,329	124	78	0.63
PRAD	prostate AC	4	57,925	585	11,702	21,769	91,981	407	48	0.12
SARC	Sarcoma	11	57,964	778	10,001	22,842	91,599	126	48	0.38
SKCM	Skin cutaneous M	9	57,964	1002	18,593	22,248	99,819	249	39	0.16
STAD	Stomach AC	7	57,964	787	18,581	26,027	103,369	295	139	0.47
UCEC	Uterine corpus EC	11	57,447	866	21,053	23,978	103,358	405	144	0.36

C. indicates carcinoma, AC Adenocarcinoma, CC Cell carcinoma, M Melanoma, and EC Endometrial carcinoma.
The third to the seventh column show the numbers of features in the respective feature groups and the eighth column the total amount of features (f). The last three columns show the numbers of observations (n), the numbers of TP53 mutation cases (m), and the ratio between the numbers of mutation events and the numbers of observations (r_m), in that order

ISSN: 1471-2105