snpXplorer: haplotype-block representation of GWAS signals for integrated variant and cross-trait interpretation

Genome-wide association studies (GWAS) have identified thousands of loci associated with complex traits, yet translating these signals into biological insight remains challenging: most associated variants are non-coding and lie within linkage disequilibrium (LD) blocks, so interpretation requires functional and cross-trait evidence assembled within haplotypic context. We present an extended version of snpXplorer, an interactive web platform that represents GWAS signals at the haplotype-block level rather than at the variant level. Haplotype-blocks are precomputed genome-wide by partitioning the genome on recombination rate and clustering variants by pairwise LD across European or African ancestries, aggregating association signals, annotations and phenotype associations onto inherited units of variation. The platform incorporates more than 11,000 European and African GWAS datasets from OpenGWAS and enables multi-scale analysis across variants, haplotype-blocks, genes and traits, integrating clinical annotations (ClinVar), allele frequencies (gnomAD), functional predictions (CADD, AlphaGenome), quantitative trait loci (GTEx), structural variation and reported GWAS associations within a single unified interface. Redundancy across closely related phenotypes is reduced by clustering trait descriptions with pretrained biomedical language embeddings, making navigation across large GWAS repositories tractable. Applied to Alzheimer's disease, snpXplorer identified a haplotype-block at the TMEM106B locus linked to eight distinct traits, revealing synergistic pleiotropy across neurological and behavioral phenotypes alongside antagonistic pleiotropy with height. The platform additionally exports PRS-ready variant sets, at user-selected p-value thresholds, for every ingested GWAS. By combining haplotype-aware representation, integrated variant annotation and cross-trait exploration, snpXplorer reduces the need for fragmented queries across databases and lowers the barrier to biological interpretation of GWAS results.