Pangenome graphs improve the analysis of rare genetic diseases

Abstract

Rare DNA alterations that cause heritable diseases are only partially resolvable by clinical next-generation sequencing due to the difficulty of detecting structural variation (SV) in all genomic contexts. Long-read, high fidelity genome sequencing (HiFi-GS) detects SVs against reference genomes with increased sensitivity and also enables the assembly of personal and graph genomes. We leveraged standard reference genomes, publicly available human haploid assemblies (n=94), together with a large collection of HiFi-GS data from a rare disease program (Genomic Answers for Kids, GA4K, n=574 assemblies). These data allowed us to build a deep population graph genome distinguishing very rare SVs from recurrent polymorphisms. Using graphs to discover SVs, we obtained a higher level of reproducibility than that obtained by the standard reference approach. We observed over 200,000 SV alleles unique to the rare disease GA4K cohort, including nearly 1,000 rare variants that impact coding sequence. With improved specificity for rare SVs, we isolate 30 candidate SVs in phenotypically prioritized genes, including known disease SVs. We isolate novel diagnostic SV in KMT2E in a patient demonstrating use of personal assemblies coupled with pangenome graphs as a new handle for rare disease genomics.