# cis-eQTL # -------- group: AA (African American) PR (Puerto Rican) MX (Mexican American) All (Pooled) AFRHp5 (AFR high) AFRLow (AFR low) IAMHp5 (IAM high) IAMLow (IAM low) files: AA.cis-eQTL.tar.gz PR.cis-eQTL.tar.gz MX.cis-eQTL.tar.gz All.cis-eQTL.tar.gz AFRHp5.cis-eQTL.tar.gz AFRLow.cis-eQTL.tar.gz IAMHp5.cis-eQTL.tar.gz IAMLow.cis-eQTL.tar.gz gene_id: gene ID from GENCODE v30 variant_id: variant ID in dbsnp150 or chr:pos:ref:alt(b38) format tss_distance: distance between variant_ID and TSS of gene ma_samples: number of samples that contains the minor allele ma_count: minor allele count maf: minor allele frequency pval_nominal: p-value slope: effect size relative to the alternative allele slope_se: standard error of the effeice size qval: per-gene qvalue (BH corection) calculated from beta-approximated p-value pval_nominal_threshold: per-gene nominal p-value threshold as described in https://github.com/francois-a/fastqtl/blob/master/doc/pages/cis_permutation.html eQTL.flag: TRUE (is an eQTL) if pval_nominal < pval_nominal_threhold vid2: variant ID in the format of chr____b38 # trans-eQTL # ---------- self-identified-race-ethnicty.out.trans.eQTL.set.freq.final.noChrX.PUB.txt.gz A union of all trans-eQTL from AA, PR, MX and pooled groups and ALT allele frequency If .P is not NA, the varaint-gene pair is a trans-eQTL-eGene pair anc-high-low.out.trans.eQTL.set.freq.final.noChrX.PUB.txt.gz A union of AFR high and IAM high filtered trans-eQTL (see definition in methods) and ALT allele frequency If AFRHp5.P or IAMHp5 is not NA, the variant-gene pair is a trans-eQTL-eGene pair in the AFR high or IAM high group geneid: gene ID from GENCODE v30 geneChr: chromosome where the gene is located vid: variant ID in the format of b38:chr:pos:ref:alt ID: variant ID in dbsnp150 or chr:pos:ref:alt(b38) format X.CHROM: chromosome where the variant is located POS: variant position REF: REF allele ALT: ALT allele geneName: gene name from GENCODE v30 .BETA: effect size relative to the ALT allele .SE: standard error of effect size .P: P value from PLINK linear regression using inverse normalized gene expression as outcome (as described in methods) .afreq: ALT allele frequency group: AA (African American) PR (Puerto Rican) MX (Mexican American) AFRHp5 (AFR high) AFRLow (AFR low) IAMHp5 (IAM high) IAMLow (IAM low) # predictdb (TWAS models) # ----------------------- TWAS models that can be used by PrediXcan or eqivalent gala.sage._Whole_Blood_tw_0.5_signif.db.weight.sorted.vcf.gz A file providing the chromosome, chromosome position, reference and alternate allele in hg38 for mapping variant ID in other formats to varaint ID used in our TWAS models ID = variant ID used in the models