eqtls

Identifier

eqtlgen.eqtls

Description

cis- and trans-eQTL meta-analysis association results across blood gene expression cohorts.

One row per (variant, gene) eQTL association tested in the meta-analysis.

Always filter on chromosome (partition column).

Filter to in_dbsnp = True and is_palindromic = False to exclude unreliable or strand-ambiguous variants before downstream analysis.

Genome-wide significance threshold is ~7.3 (-log10(5e-8)); use false_discovery_rate or neg_log_10_p_value_bonferroni for stricter multiple-testing control.

Related tables: eqtlgen.variants (join on variant_id, chromosome); eqtlgen.genes (join on gene_id); dbsnp.vcf (join on rsid, chromosome); open_targets.targets (join on gene_id).

Partitions

• location
  • 'cis'
  • 'trans'
• chromosome
  • '1'
  • '10'
  • '11'
  • '12'
  • '13'
  • '14'
  • '15'
  • '16'
  • '17'
  • '18'
  • '19'
  • '2'
  • '20'
  • '21'
  • '22'
  • '3'
  • '4'
  • '5'
  • '6'
  • '7'
  • '8'
  • '9'

Schema

Column	Type	Description
gene_id	TEXT	Ensembl gene ID for the eQTL transcript.
location	TEXT	Partition column. eQTL class — ‘cis’ (variant near the gene) or ‘trans’ (distant or different chromosome). Partition column.
chromosome	TEXT	Partition column. Chromosome on which the variant is located.
position_grch37	INT	Variant position in assembly GRCh37 coordinates.
position_grch38	INT	Variant position in assembly GRCh38 coordinates.
effect_allele	TEXT	Effect allele. Harmonized to equal the dbSNP alt_allele.
other_allele	TEXT	Non-effect allele. Harmonized to equal the dbSNP ref_allele.
rsid	TEXT	dbSNP rsID (e.g. rs123456). May be null for novel variants.
variant_id	TEXT	Unique variant identifier within the study. Format is chromosome:position_grch37:AlleleA:AlleleB, where AlleleA/AlleleB are the source-file alleles before harmonization.
neg_log_10_p_value	FLOAT	-log10(p-value). Higher = more significant.
neg_log_10_p_value_bonferroni	FLOAT	-log10(Bonferroni-corrected p-value).
z	FLOAT	Z-score of the eQTL effect, oriented to effect_allele after harmonization (sign flipped where the source allele assignment was swapped).
n	INT	Sample size contributing to the meta-analysis.
n_cohorts	INT	Number of cohorts contributing to the meta-analysis.
false_discovery_rate	FLOAT	Benjamini-Hochberg FDR.
dbsnp_build	TEXT	dbSNP build version that the variant was harmonized against (e.g. b157). Null for variants without a dbSNP match.
in_dbsnp	BOOLEAN	Variant has a match in dbSNP. Filter to in_dbsnp = True to exclude rows whose effect-allele orientation could not be validated against a reference.
is_palindromic	BOOLEAN	Variant alleles are A/T or C/G. These require specific treatment during harmonization and are typically excluded from cross-cohort analyses.
was_swapped	BOOLEAN	Effect and other alleles in the source files were in the opposite order from dbSNP REF/ALT (on the aligned strand). When true, the Z-score sign and allele-frequency fields were inverted during harmonization.
was_flipped	BOOLEAN	Source files reported the variant on the reverse strand relative to dbSNP; alleles were reverse-complemented during harmonization. Null for variants without a confident strand assignment.