A global assessment of cancer genomic alterations in epigenetic mechanisms

Table 1 Most frequently mutated or over/under-expressed chromatin factors in cancer

Gene	Cancer	Mutated (%)	q value	Gene	Cancer	Overexpressed (%)	Rank (out of 20,544)	Gene	Cancer	Underexpressed (%)	Rank (out of 20,544)
ATRX	LGG	39.8	1.82E-11	HJURP	PRAD	88	3	PRDM16	KIRP	100	14
ARID1A	UCEC	32.3	8.11E-12	UHRF1	BRCA	94.7368	7	PRDM16	KIRC	95.8904	40
PBRM1	KIRC	28.5	1.76E-11	TDRD1	PRAD	84	14	KAT2B	HNSC	82.9268	55
DNMT3A	LAML	25.4	3.96E-12	DPF1	HNSC	87.8049	25	JMJD5	LIHC	88	56
MLL2	LUSC	19.1	6.39E-05	HJURP	BRCA	91.2281	29	TDRD10	LUSC	100	57
ARID1A	STAD	18.1	2.63E-11	HJURP	LUAD	94.7368	31	MECOM	KIRC	95.8904	62
MLL2	HNSC	17.7	1.94E-11	HJURP	LIHC	96	34	L3MBTL4	PRAD	76	73
CHD4	UCEC	13.3	0.000958	ASF1B	BRCA	91.2281	37	ASXL3	THCA	78.7879	81
STK31	SKCM	11.9	0.049702	UHRF1	KIRP	90	37	CECR2	THCA	80.303	83
NSD1	HNSC	10.8	1.89E-11	ORC1	LIHC	94	40	ZCWPW2	HNSC	80.4878	84
SETD2	KIRC	9.8	1.76E-11	EZH2	LIHC	94	44	ZGPAT	LIHC	82	141
TET2	LAML	8.6	2.42E-11	ASF1B	KIRP	90	45	TDRD10	BRCA	85.0877	154
EP300	HNSC	7.8	0.008196	HJURP	LUSC	100	47	MECOM	KIRP	93.3333	162
MLL3	BRCA	7.7	1.95E-11	UHRF1	LUSC	100	68	ACTA1	PRAD	70	194
SMARCA4	LUAD	7.4	2.99E-08	ORC1	LUAD	92.9825	72	CBX7	BRCA	83.3333	218
SETD2	LUAD	7	0.006858	ORC1	LUSC	100	79	TDRD9	THCA	71.2121	261
KDM5C	KIRC	5.8	3.75E-11	ASF1B	LUSC	100	90	TDRD6	BRCA	81.5789	277
ARID1A	LUAD	5.7	0.001823	EZH2	LUSC	100	94	CHD5	PRAD	68	277
SMARCA4	LGG	4.5	0.035502	UHRF1	LUAD	89.4737	101	KIAA1045	BRCA	80.7018	298
KDM5C	LUAD	4.4	0.091808	HIST1H3D	BRCA	85.9649	102	CBX7	LUAD	85.9649	342
ARID1A	BRCA	2.9	0.000117	TCF19	LIHC	88	102	SCML2	THCA	69.697	344
PHF6	LAML	2.5	0.001759	HJURP	KIRP	83.3333	107	L3MBTL4	BRCA	79.8246	346
HIST1H1B	HNSC	2.3	0.084637	CECR2	PRAD	72	107	DPF3	LIHC	72	364
HIST1H3B	BRCA	1.1	9.81E-05	HIST1H3H	BRCA	85.9649	109	CECR2	KIRC	83.5616	388
				UHRF1	LIHC	88	109	PRDM16	LUSC	92	410

Values indicate the frequency of cancer patients where a gene is mutated (non-silent mutations only), where Log2(mRNA tumor/matched control) >1 for overexpression, or < -1 for underexpression. All data were extracted from TCGA and the ICGC. Patient cohorts are greater than 30 for overexpression and 100 for mutations. Hypermutated genomes and other sources of noise were excluded (as detailed in the Methods section). q value is a measure of the P value adjusted for false discovery rate. Ranks indicate the number of genes more frequently over- or underexpressed in the human genome.

ISSN: 1756-8935