Building and applying pangenome references to capture genetic diversity.

Reference genomes serve as a coordinate system and are central to almost all analyses in genomics. However, linear reference genomes are based on a single individual or a small number of individuals and do not represent genetic diversity. Recent advances in de novo genome assembly, powered by long-read sequencing technologies, now enable the sequence reconstruction of many genomes to reference quality. These pangenomes integrate sequences from multiple individuals into graph-based or multi-haplotype representations, capturing genetic variation beyond a single linear reference. The widespread adoption of such pangenome references, which encode a diverse set of haplotypes, thus removes biases and enables the discovery of variants relative to all included haplotype backgrounds. The emergence of corresponding computational tools for analysing structural variants and complex genetic loci opens up opportunities in genome-wide association studies and rare-disease genetics. Here we review these opportunities, as well as challenges concerning pangenomes that need to be addressed by the research community.