Enter an interval, or select a gene or candidate record.
Glossary and evidence definitions
Terms used in the overlap, pairwise and integrated browser results.
Overlap analysis
GRB
Genomic regulatory block: a conserved long-range regulatory
domain containing a predicted target gene and neighbouring
bystander genes.
overlapEpiGene
A gene–trait–chromosome set with at least two qualifying,
mutually non-overlapping GWAS LD blocks assigned to the gene.
overlapEpiEnh
An overlapEpiGene set in which a significantly associated
FANTOM5 enhancer physically overlaps at least one qualifying LD
block.
Gene–trait set
One gene, mapped GWAS trait and chromosome considered as a
single counting unit.
Overlap row
One enhancer–gene–trait–LD-block association. Several rows
may describe the same entity under different traits or blocks.
LD in GRB
The LD-block interval physically overlaps the GRB assigned
to the corresponding gene. This differs from the gene merely
having a GRB assignment.
Gene roles
GRB-trgt
Predicted long-range regulatory target gene within a GRB.
GRB-byst
Bystander gene located within the target gene’s GRB.
GWAS-trgt / GWAS-byst
GRB target or bystander gene with additional GWAS
annotation.
GWAS-out
GWAS-associated gene outside the GRB target and bystander
categories.
Pairwise non-additivity
Tier 1 pair
The gene passes the baseline-linear versus pairwise
comparison (p < 0.01, ΔBIC < −10), and the pair
has permutation FDR < 0.05 with significance in at least
three of the four pair-level tests. In Candidate evidence,
Tier 1 means that the enhancer participates in such a pair for
the displayed gene.
Permutation FDR
BH-adjusted empirical significance of the enhancer product
term relative to randomized tissue alignment; this is the
strictest individual pair-level test.
Revised RC FDR
BH-adjusted residual-product correlation test after
accounting for the baseline-adjusted additive model.
Best FDR
The smallest FDR among the available pair-level tests,
including the revised residual-correlation test.
Significant tests
Number of tests significant at FDR < 0.05, including the
revised residual-correlation result in the displayed count.
Interaction score
A ranking aid combining −log10(best FDR) with the number of
significant tests. It is not an additional hypothesis test.
Direction
Consensus direction of the non-additive signal across
compatible tests; it is descriptive rather than a separate
significance criterion.
Candidate rank
Ordering within the conservative candidate list; it does
not replace the individual statistics.
Integration between analyses
Exact gene–enhancer support
The same gene and enhancer coordinate occur in both the
overlap and pairwise analyses.
Cross-support
The strongest correspondence between analyses: exact Tier
1, exact any-test FDR < 0.05, exact tested pair,
enhancer-coordinate only, gene only, overlap-only, or
pairwise-only.
Enhancer-coordinate only
The enhancer occurs in both analyses, but not with the same
gene assignment.
Gene only
The gene occurs in both analyses without an exact shared
enhancer coordinate.