A pan-cancer analysis of driver gene mutations, DNA methylation and gene expressions reveals that chromatin remodeling is a major mechanism inducing global changes in cancer epigenomes

Table 1 The identified 32 MDGs

MDGs	\|A_i\|	\|T_i\|	P_i	D_i	T^-_i	T⁺_i
TP53	15	8	< e-06	both^a	BLCA BRCA HNSC LIHC LUAD STAD UCEC	LGG^a
PTEN	14	3	0.00147	both	LGG	STAD UCEC
RB1	11	2	0.00861	both	LGG	BLCA
PIK3CA	11	1	0.0277	hyper		STAD
ARID1A	10	1	0.0401	hyper		STAD
KRAS	8	1	1.2e-05	hypo	TGCT
KMT2D	8	2	0.00575	both	BLCA	STAD
NF1	6	2	0.000227	hypo	LGG PCPG
CTNNB1	6	2	0.00185	hypo	LIHC UCEC
SETD2	5	2	0.00351	hyper		KIRC KIRP
KMT2C ^b	5	1	0.00681	hyper		STAD
EGFR	4	1	9.5e-05	hypo	LGG
HRAS	4	2	0.000266	both	PCPG	HNSC
BRAF	4	2	0.00474	both	THCA	COAD
IDH1	3	2	< e-06	hyper		GBM LGG
CIC	3	1	1.3e-05	hyper		LGG
NRAS	3	2	1.4e-05	both	TGCT	THCA
RNF43	3	2	5.1e-05	both	STAD	COAD
ATRX	3	1	0.000487	hyper^a		LGG^a
ZBTB20	3	2	0.00151	hyper		COAD STAD
NOTCH1	3	1	0.00189	hyper		LGG
CDH1	3	2	0.00213	hyper		BRCA STAD
KEAP1	3	1	0.00578	hypo	LUAD
SMARCA4	3	1	0.00687	hypo	LUAD
FOXA1 ^b	3	1	0.00953	hypo	PRAD
EPHA2 ^b	3	1	0.0104	hyper		STAD
KIT	2	1	< e-06	hypo	TGCT
KMT2B	2	2	0.000511	both	STAD	COAD
FGFR3 ^b	2	1	0.000697	hypo	BLCA
STK11 ^b	2	1	0.00193	hypo	LUAD
NSD1	1	1	< e-06	hypo	HNSC
BAP1 ^b	1	1	0.000142	hyper		LIHC

|Ai|: number of tumor types in which CDG i is mutated in ≥ 5 samples with available methylation data;
|Ti|: number of tumor types whose genome-wide methylation levels are significantly associated with the mutation status of CDG i;
p_i: p-value testing if CDG i is significantly associated with genome-wide methylation changes across tumor types;
Di: direction of methylation changes associated with mutation status of CDG i;
T⁺_i: tumor types that are hyper-methylated by CDG i;
T^-_i: tumor types that are hypo-methylated by CDG i;
^a: Further stratified analysis by IDH1 mutation status in LGG tumor samples suggests an opposite direction from hyper- to hypo-methylation;
^b: genes that are not overlapping driver genes
__ : genes that are identified as associated with genome-wide patterns of aberrant methylation by Chen et al. [5]

ISSN: 1755-8794