Oral vitamin D supplementation induces transcriptomic changes in rectal mucosa that are linked to anti-tumour effects

Table 2 Gene-set testing for enrichment of the candidate gene-set from Phase 1 after supplementation in Phase 2 and the BEST-D study

PHASE 1 Correlative dataset	SCOVIDS Phase 2				BEST-D trial
	Intervention dataset, NM All participants		Intervention dataset, blood, HT12		Blood, HT12
	HT12	RNA-seq	All participants	With vs. without rectal response	All participants
Candidate gene-set: positive association with 25-OHD level	P < 1.0E−07 n = 349	2.05E−07 n = 242	P = 3.89E−13 n = 239	P = 3.65E−12 n = 291	P = 4.72E−06 n = 191
Candidate gene-set: negative association with 25-OHD level	P = 2.8E−05 n = 206	6.87E−09 n = 155	P = 9.60E−05 n = 185	P = 3.50E−17 n = 217	P = 0.02 n = 130

NM normal mucosa. P value given for directional gene-set enrichment test of whether Phase 1 candidate gene-set showed greater change after supplementation when compared to randomly chosen genes. n = number of genes tested. Rectal response is defined by candidate gene-set enrichment in NM P < 0.001. To explore the role of VDR genotype in modifying the association between 25-OHD and gene expression, VDR haplotypes were derived for the four genotyped VDR polymorphisms, rs1544410, rs10735810, rs7975232, rs11568820, using BEAGLE software (version 3.3.2) with standard settings [59] and used as additional covariates in the linear regression model in a subset of 125 participants, resulting in a smaller candidate gene-set which remained significantly enriched after supplementation (P < 0.0001)

ISSN: 1741-7015