Pan-cancer analysis of pre-diagnostic blood metabolite concentrations in the European Prospective Investigation into Cancer and Nutrition

Table 4 Other associations identified in a large proportion of the bootstrap samples. Associations identified in at least 55% of both bootstrap analyses are reported, along with the proportion of bootstrap samples in which they were identified, and the corresponding average log-odds ratio (as estimated by the data-shared lasso on each bootstrap sample)

Feature	Cancer type^a	Proportion of bootstrap samples¹	Average log-OR^b	Proportion of bootstrap samples²	Average log-OR^c
Overall associations with cancer risk
Glutamate	Overall	--	--	55%	0.09
Spermine	Overall	--	--	78%	−0.10
Type-specific associations
gly_Clus	EnC	65%	−0.17	78%	−0.14
c10_1_Clus	KiC	56%	−0.18	56%	−0.17
lysopc_a_c16_1	Loc.PrC	84%	−0.19	78%	−0.18
Arginine	KiC	74%	0.23	71%	0.21
lysopc_a_c16_0_Clus	Loc.PrC	86%	0.24	79%	0.22
Glutamate	BrC	--	--	56%	−0.14
Serotonin	CRC	--	--	84%	0.35

^aBrC stands for breast cancer, CRC for colorectal cancer, EnC for endometrial cancer, KiC for kidney cancer, and Loc.PrC for localized prostate cancer
^bBootstrap samples were generated from the original sample of 5828 matched case-control pairs with information on 117 metabolites (corresponding to 50 features after the clustering step)
^cBootstrap samples were generated from the original sample which comprised 4761 matched case-control pairs with information on 133 metabolites (corresponding to 65 features after the clustering step) after excluding the participants of the second CRC study

ISSN: 1741-7015