Exploring quantitative traits-associated copy number deletions through reanalysis of UK10K consortium whole genome sequencing cohorts

Abstract

Objectives We performed comprehensive association analyses of common high-confidence gnomAD-reported copy number deletions (CNDs) with 60 quantitative traits from UK10K consortium WGS data.

Methods The study made use of data generated by the UK10K Consortium. UK10K consortium WGS data consist of TwinsUK (n = 1754, middle-aged females) and ALSPAC (n = 1867, birth to adolescence) cohorts. UK10K consortium called 18,739 CNDs (hg19) with GenomeSTRiP software. After filtering out variants with minor allele frequency < 0.05 or HWE P < 1.0 × 10− 6, 1222 (TwinsUK) and 1211 (ALSPAC) CNDs remained for association analyses with 60 normalized quantitative traits.

Results We identified 23 genome-wide significant associations at 13 loci, among which 2 associations reached experiment-wide significance. We found that two common deletions in chromosome 4, located between WDR1 and ZNF518B (23.3 kb, dbVar ID:nssv15888957, 4:10211262–10,234,569 and 9.8 kb, dbVar ID:nssv15888975, 4:10392422–10,402,191), were associated with uric acid levels (P = 5.23 × 10− 11 and 2.29 × 10− 8, respectively). We also discovered a novel deletion spanning chromosome 18 (823 bp, dbVar ID: nssv15841628, 8:74347187–74,348,010) associated with low HDL cholesterol levels (P = 4.15 × 10− 7). Additionally, we observed two red blood cell traits-associated loci with genome-wide significance, a 13.2 kb deletion in 7q22.1 (nssv15922542) and a 3.7 kb deletion in 12q24.12 (nssv15813226), both of which were located in regions previously reported to be associated with red blood cell traits. Two deletions in 11q11 (nssv15803200 and nssv15802240), where clusters of multiple olfactory receptor genes exist, and a deletion (nssv15929560) upstream to DOCK5 were associated with childhood obesity. Finally, when defining Trait-Associated copy number Deletions (TADs) as CNDs with phenotype associations at sub-threshold significance (P < 10− 3), we identified 157 (97.5%) out of 161 TADs in non-coding regions, with a mean size of 4 kb (range: 209 − 47,942 bp).

Conclusion We conducted a reanalysis of the UK10K Whole Genome Sequencing cohort, which led to the identification of multiple high confidence copy number deletions associated with quantitative traits. These deletions have standard dbVar IDs and replicate previous findings, as well as reveal novel loci that require further replication studies.