An optional feature in eQTLA can identify markers in close proximity that are highly correlated due to apparent linkage disequilibrium (LD). This is accomplished by pairwise calculation of either D’ or r2 values between markers within a given genomic interval. A C++ implementation of Gregory Warnes’s LD function from the R package genetics is used for this operation. Each group of markers in high LD is assigned a unique locus ID. Post-hoc adjustment can be applied to eQTL data by amalgamating markers from the same locus and assigning the median log10 converted P value to the cluster.
