Summary level-data accompanying "A functional locus at 8q21.13 associated to FABP4 levels and causally links coronary artery disease and type 2 diabetes"

Introduction

These are the Summary Level-data as presented in:

"A functional locus at 8q21.13 associated to FABP4 levels and causally links coronary artery disease and type 2 diabetes". Unpublished. (tentative title)

If you use these data please cite this DOI or the (pre)print when available. When you have any questions or comments regarding this study or these files, please contact me via:

Sander W. van der Laan, PhD | Central Diagnostics Laboratory, Division Laboratory, Pharmacy and Biomedical genetics, Circulatory Health Program, University Medical Center Utrecht, Utrecht University | s.w.vanderlaan-2 [at] umcutrecht [dot] nl or s.w.vanderlaan [at] gmail [dot] com | @swvanderlaan

Files and description

There are three files available:

1. meta.GWAS.FABP4.1Gp1.EUR.MODEL1.* - Gzipped file containing all the (unfiltered) meta-analysis results for model 1 (FABP4 ~ SNP + age + sex + PC1-10 + study specific covariates).
   • Discovery dataset ends with: "summaryQC.ELISAonly.txt.gz"
   • Replication dataset ends with: "summaryQC.OLINKonly.txt.gz"
   • Combined dataset ends with: "summaryQC.txt.gz"
2. meta.GWAS.FABP4.1Gp1.EUR.MODEL2.* - Gzipped file containing all the (unfiltered) meta-analysis results for model 2 (FABP4 ~ SNP + age + sex + PC1-10 + study specific covariates + BMI).
   • Discovery dataset ends with: "summaryQC.ELISAonly.txt.gz"
   • Replication dataset ends with: "summaryQC.OLINKonly.txt.gz"
   • Combined dataset ends with: "summaryQC.txt.gz"
3. meta.GWAS.FABP4.1Gp1.EUR.MODEL3.* - Gzipped file containing all the (unfiltered) meta-analysis results for model 3 (FABP4 ~ SNP + age + sex + PC1-10 + study specific covariates + BMI + eGFR).
   • Discovery dataset ends with: "summaryQC.ELISAonly.txt.gz"
   • Replication dataset ends with: "summaryQC.OLINKonly.txt.gz"
   • Combined dataset ends with: "summaryQC.txt.gz"

All these files have the same lay-out and are gzipped. The reference used for meta-analysis of GWAS was 1000G phase 1, version 3 (so called 'ALL.wgs.integrated_phase1_v3.20101123.snps_indels_sv.sites'-panel) using data from the EUR populations. For a more detailed explanation of these columns and the reason to include them, please refer to De Bakker et al. Hum Mol Genet 2008. 

• VARIANTID - variantID as represented in 1000G phase 1, version 3.
• CHR - chromosome numbers [1-22 and X, Y, MT].
• POS - base pair position.
• MINOR - minor allele as present in 1000G.
• MAJOR - major allele as present in 1000G.
• MAF - minor allele frequency as present in 1000G.
• CODEDALLELE - coded allele, i.e. the effect allele, as represented (and harmonized) across cohorts. Note that this is not necessarily the minor allele!
• OTHERALLELE - the other allele, i.e. the non-effect allele.
• CAF - coded allele frequency, i.e. the effect allele frequency. Note that this is not necessarily the minor allele frequency!
• N_EFF - the effective sample size corrected for the imputation quality.
• Z_SQRTN - Z-score of the effective sample-size-weighted meta-analysis.
• P_SQRTN - P-value of the effective sample-size-weighted meta-analysis.
• BETA_FIXED - beta from the fixed-effects model.
• SE_FIXED - standard error from the fixed-effects model.
• Z_FIXED - z-score from the fixed-effects model.
• P_FIXED - P-value from the fixed-effects model.
• BETA_LOWER_FIXED - 95% lower confidence interval of the beta from the fixed-effects model.
• BETA_UPPER_FIXED - 95% upper confidence interval of the beta from the fixed-effects model.
• BETA_GC - beta after correcting the fixed-effects beta for genomic inflation.
• SE_GC - standard error after correcting the fixed-effects SE for genomic inflation.
• Z_GC - Z-score after correcting the fixed-effects Z-score for genomic inflation.
• P_GC - P-value after correcting the fixed-effects p-value for genomic inflation.
• BETA_RANDOM - beta from the random-effects model.
• SE_RANDOM - standard error from the random-effects model.
• Z_RANDOM - Z-score from the random-effects model.
• P_RANDOM - P-value from the random-effects model.
• BETA_LOWER_RANDOM - 95% lower confidence interval of the beta from the random-effects model.
• BETA_UPPER_RANDOM - 95% upper confidence interval of the beta from the random-effects model.
• COCHRANS_Q - Cochran's Q as a measure of heterogeneity between studies (see this wiki).
• DF - degrees of freedom, equals the number of studies included for the respective variant (N) minus 1, i.e. DF = N-1.
• P_COCHRANS_Q - P-value of Cochran's heterogeneity test.
• I_SQUARED - I² as a measure of heterogeneity between studies (see this wiki).
• TAU_SQUARED - Tau² as a measure of true heterogeneity between studies (see this wiki).
• DIRECTIONS - the sign of beta in each contributing cohort, annotated as “.” if the variant is missing from a particular cohort.
• GENES_250KB - list of all genes as mapped using GENCODE v19 (GRCh37, hg19, Feb2009) with 250kb.
• NEAREST_GENE - the gene closest to the respective variant.
• NEAREST_GENE_ENSEMBLID - the ENSEMBLID of the nearest gene.
• NEAREST_GENE_STRAND - strand on which the nearest gene is present.
• VARIANT_FUNCTION - variant function as taken from dbSNP v150.
• CAVEAT - potential issue as reported by MetaGWASToolKit, e.g. if the variant is an A/T or C/G SNP with allele frequency between 0.35 and 0.65 (indicating strandedness ambiguity).
• QC - utility column, can be used to filter out all the variants with e.g. 'CAF' < 0.001, 'DF' <= 2, 'N_EFF' < 5000 and 'CAVEAT' having an issue depending on the dataset used (discovery, replication, or combined).