Pan-cancer stratification of solid human epithelial tumors and cancer cell lines reveals commonalities and tissue-specific features of the CpG island methylator phenotype

Table 1 Cancer types, sample sizes and probe-set cardinalities

Cancer type	Variably methylated probes	Differentially methylated probes	Control	Tumor	CIMP-	CIMPi	CIMP+	CIMP−	CIMP+
Cancer type	Variably methylated probes	Differentially methylated probes	Control	Tumor	CIMP-	CIMPi	CIMP+	pan-cancer	pan-cancer
BLCA (bladder urothelial carcinoma)	49,148	338	20	201	78	84	39	43	14
BRCA (breast invasive carcinoma)	46,722	1,311	96	676	270	244	162	76	47
COAD (colon adenocarcinoma)	46,168	2,656	38	274	96	92	86	71	60
HNSC (head and neck squamous cell carcinoma)	44,100	1,228	50	426	156	186	84	115	55
KIRC (kidney renal clear cell carcinoma)	26,148	196	160	296	126	94	76	97	65
KIRP (kidney renal papillary cell carcinoma)	28,083	40	45	147	60	59	28	NA	NA
LIHC (liver hepatocellular carcinoma)	51,875	544	50	151	45	61	45	NA	NA
LUAD (lung adenocarcinoma)	42,822	1,667	32	437	161	169	107	67	48
LUSC (lung squamous cell carcinoma)	40,606	1,430	42	359	140	142	77	32	11
PAAD (pancreatic adenocarcinoma)	27,899	1,602	9	65	16	33	16	NA	NA
PRAD (prostate adenocarcinoma)	33,718	450	49	248	74	122	52	NA	NA
READ (rectum adenocarcinoma)	40,496	1,255	7	96	31	39	26	22	22
STAD (stomach adenocarcinoma)	62,606	1,110	2	260	109	95	56	NA	NA
THCA (thyroid carcinoma)	21,945	0	56	508	NA	NA	NA	NA	NA
UCEC (uterine corpus endometrioid carcinoma)	43,040	1,430	46	407	155	139	113	54	34

Probe set cardinalities and sample sizes for the 15 cancer types that were included in our analysis. The last two columns show the number of CIMP+ and CIMP− samples from our genome-wide methylation study that also appear in the selected functional event data matrix from Ciriello et al. [27].

ISSN: 1756-8935