These files include summary results for the diet genome-wide association analysis (Merino J, et al. Nature Human Behavior 2022; PMID: 34426670). We conducted single trait genetic association analysis for carbohydrate, fat, and protein preference and multivariate genome-wide association analysis in 282,271 participants of European ancestry from the UK Biobank (n = 191,157) and Cohorts for Heart and Aging Research in Genomic Epidemiology Consortium (n = 91,114). Results are presented with or without BMI adjustment. Further information and file format details are provided below.
 
Brief methods: 
The proportion of total energy intake from carbohydrate, fat, and protein was calculated for each participant. Models were adjusted for age, sex, 20 principal components of ancestry, and genotyping array and accounted for sample relatedness. A sensitivity analysis was conducted after adjusting for BMI. We performed a fixed-effects inverse-variance weighted meta-analysis using METAL, and applied double genomic control (GC). We used a minor allele count threshold of 20 to select the genetic variants, and variants with low imputation quality (<0.4) were removed. Single-trait meta-analyses estimates for carbohydrate, fat, and protein were combined in a multivariate analysis using CPASSOC.

For each single-trait dataset variant, we have provided the following information: 
- MarkerName: the chromosome and the base pair location the variant is located on based in build 37, and the reference allele and the alternate allele on the fwd strand.
- Allele1: the effect allele of the variant.
- Allele2: the non-effect allele.
- Freq1: frequency of the effect allele in the meta-analysis
- FreqSE: standard error of the frequency of the effect allele in the meta-analysis
- MinFreq: minimum frequency of the effect allele in the meta-analysis
- MaxFreq: maximum frequency of the effect allele in the meta-analysis
- Effect: the effect size of the effect allele in the meta-analysis
- stdErr: the standard error of the effect size in the meta-analysis
- P-value: the p-value of the variant in the meta-analysis
- Direction: direction of the estimated effect size from each study contributing in the meta-analysis
- HetISq: I2 heterogeneity statistic
- HetChiSq: heterogeneity test statistic
- HetDf: degrees of freedom for heterogeneity test
- HetPVal: heterogeneity test p-value


For each variant in the multivariate genome-wide association analysis, we have provided the following information:
- SNP: the chromosome and the base pair location the variant is based in build 37, and the reference allele and the alternate allele on the fwd strand.
- A1: the effect allele of the variant.
- carb.p: the p-value of the variant in the carbohydrate meta-analysis
- fat.p: the p-value of the variant in the fat meta-analysis
- pro.p: the p-value of the variant in the protein meta-analysis
- sHom.chisq: sHom heterogeneity test statistic
- sHom.p: sHom p-value
- SHet.chisq: sHet heterogeneity test statistic
- SHet.p: sHet p-value