Abstract
ABSTRACT
Research on genetic variants affecting the gut microbiome and blood metabolites has mainly focused on single nucleotide variants (SNPs), leaving structural variants (SVs) largely unexplored. Here, we analyzed high‐coverage (∼42×) whole‐genome sequencing data from 2002 individuals and constructed a catalog of 138 859 high‐confidence SVs, including 44.3% variants not represented in public SV resources. Genome‐wide association analyses of 22 519 common SVs identified 30 microbiome‐associated and 48 metabolite‐associated SV loci, including 19 study‐wide significant associations. Notable signals included an Asian‐specific 6.5k deletion at
TENT2
associated with
Coprobacillus sp. 29_1
abundance (β = 0.158,
p
= 2.5 × 10
−9
) and a deletion covering
HBA1/HBA2/HBM/HBQ1
linked to nine blood metabolic traits (strongest for MCV; β = ‐0.570,
p
= 1.22 × 10
−91
). Conditional and variance‐partitioning analyses indicated that SVs capture complementary genetic information beyond tested SNP signals. SV‐based Mendelian randomization delineated putative directional relationships, including
Clostridium ramosum
with serum uric acid and hyperuricemia. In vitro experiments further supported three predicted metabolite–microbe interactions, showing that vitamin K1 inhibited
Coprococcus comes
, hydroxyproline inhibited
Veillonella parvula
, and cysteine promoted
Fusobacterium mortiferum
growth. Together, our study provides a comprehensive SVs resource and highlights SVs as overlooked genetic determinants of host‐microbiome‐metabolome interactions relevant to human physiology and disease.