We carried out GWAS analysis (imputed to 1000 Genomes Phase 3) for 136 metabolites (14 amino acids, 42 acylcarnitines, and 80 sphingolipids) in up to 1,954 Singapore Chinese. These individuals came from the Singapore Prospective Study Program (SP2), and were genotyped on three separate Illumina arrays. Each variant was tested for association with metabolites in two regression models. In the first model to obtain P-values, we used inversed normalised residuals after regressing out age, age2, sex, and BMI to test for association with each variant. The second model regressed log2 transformed metabolite with age, age2, sex, BMI and genetic dosage for interpretable effect size. We then carried out fixed-effects sample-size weighted meta-analysis as implemented in METAL meta-analysis tool to combine association statistics from the three arrays.
The tarball file contains meta-analysis summary statistics for the 136 metabolites analysed in this study. We reported variants with association statistics in at least two arrays.
Each file contains the following information:
- CHR: chromosome
- POS: position (hg19)
- MARKERNAME: RSID
- N: Total sample size
- EFF_ALLELE: Effect allele
- NONEFF_ALLELE: Other allele
- EFF_ALLELE_FREQ: Effect allele frequency
- BETA: Effect estimate from analysis of log2 transformed metabolites
- SEBETA: Standard error of effect estimate
- ZSCORE: Combined z-statistics for the variant from sample-size weighted meta-analysis
- P: P-value
Reference: Chai JF, et al. Associations with metabolites in Chinese suggest new metabolic roles in Alzheimer’s and Parkinson’s diseases. Hum Mol Genet. 2019 Oct 19. pii: ddz246. doi: 10.1093/hmg/ddz246. PubMed PMID: 31628463.
Please refer any queries to Dr Xueling Sim (firstname.lastname@example.org)