Distinguishing potential bacteria-tumor associations from contamination in a secondary data analysis of public cancer genome sequence data

Table 3 Relative Mycobacterium spp. read pairs and sequencing center for each sample

Cancer type	Sequencing centers
	Total samples¹					Samples with bacteria					Samples with a log₁₀ (Myco./total reads) ratio ≥ −7					Samples with a log₁₀ (Myco./total reads) ratio ≥ −4
	Baylor²	BCGSC	Broad	UNC	Wash.	Baylor	BCGSC	Broad	UNC	Wash.	Baylor	BCGSC	Broad	UNC	Wash.	Baylor	BCGSC	Broad	UNC	Wash.
AML	–³	192	–	–	–	–	117	–	–	–	–	33	–	–	–	–	0	–	–	–
BRCA	–	340	–	393	–	–	0	–	90	–	–	0	–	40	–	–	0	–	0	–
GBM	1	–	154	–	3	0	–	125	–	3	0	–	29	–	1	0	–	0	–	0
KIRC	–	397	–	248	–	–	0	–	72	–	–	0	–	26	–	–	0	–	0	–
KIRP	–	–	–	15	–	–	–	–	9	–	–	–	–	2	–	–	–	–	0	–
LUAD	–	58	–	18	–	–	0	–	10	–	–	0	–	3	–	–	0	–	0	–
LUSC	–	136	–	167	–	–	0	–	20	–	–	0	–	4	–	–	0	–	0	–
OV	3	–	262	–	97	0	–	197	–	20	0	–	106	–	0	0	–	1	–	0
STAD	–	71	–	–	–	–	71	–	–	–	–	22	–	–	–	–	0	–	–	–

¹ Total samples and samples with bacteria are defined as an analyte from each sample, so if one tumor sample had DNA and RNA sequenced or had RNA sequenced at multiple centers, each analysis of each nucleic acid is counted individually. For example, while there were only 451 BRCA samples (Table 1), most samples were sequenced at two sequencing centers, resulting in a total of 733 entries in the “Total samples” column of this table.
²Baylor = Baylor College of Medicine Human Genome Sequencing Center; BCGSC = British Columbia Genome Sequencing Center; Broad = The Broad Institute; UNC = University of North Carolina, Chapel Hill; Wash. = Washington University in St. Louis.
³Sequencing centers that did not sequence any samples of a specific cancer type are denoted by a "–"

ISSN: 2049-2618