Characterization and evaluation of gene fusions as a measure of genetic instability and disease prognosis in prostate cancer

Table 1 Numbers of detected gene fusions

		FF_RP control	FF_RP tumor	TCGA_PRAD tumor	FFPE_Bx tumor
	number of samples	24	40	332	176
all confidences	number of fusions	633	2,373	12,908	109,590
	average fusions per sample	26.38	59.33	38.88	622.67
	number of unique fusions per cohort	519	1,952	8,790	109,281
high	number of fusions (percentage of all confidences)	34 (5.37%)	488 (20.56%)	3,265 (25.29%)	230 (0.21%)
	mean fusions per sample	1.42	12.20	9.83	1.31
	number of unique fusions per cohort	31	447	3,072	204
medium	number of fusions (percentage of all confidences)	209 (33.02%)	752 (31.69%)	2,134 (16.53%)	7,476 (6.82%)
	mean fusions per sample	8.71	18.80	6.43	42.48
	number of unique fusions per cohort	159	567	1,778	7,440
low	number of fusions (percentage of all confidences)	415 (65.56%)	1,273 (53.65%)	8,002 (61.99%)	101,946 (93.02%)
	mean fusions per sample	17.29	31.83	24.10	579.24
	number of unique fusions per cohort	350	1,065	4,393	101,693

The columns depict the number of detected gene fusions per cohort. We report the numbers of control and tumor-free samples of FF_RP (column 1), the tumor samples of FF_RP (column 2), all samples of TCGA_PRAD (column 3), and all samples of FFPE_Bx (column 4). The rows depict the number of detected fusions, the corresponding mean of fusions per sample, and the unique occurrence of fusions across a cohort. The numbers are listed for the whole cohort (‘all confidences’) and separated for each Arriba confidence level (‘high’, ‘medium’, and ‘low’). Percentages in parentheses correspond to the proportion of each category compared to the total number of fusions per cohort. Percentages do not sum up to 100% because fusions can appear in multiple confidence groups

ISSN: 1471-2407