Building tools and community to make pangenomes accessible

This $1,645,076 Cooperative Agreement award from the National Human Genome Research Institute (NHGRI) under the Human Genome Research (CFDA 93.172) program will develop advanced tools and resources to enable new ways to access and understand the human pangenome. The key products and services to be delivered include:

• Sequence alignment and population genetic analysis methods to rapidly search, compare, and interpret pangenome data, leveraging succinct data structures for scalable pangenome indexing and analysis.
• Interactive visualizations using GPU-accelerated algorithms to provide an intuitive, human-scale interface for exploring pangenome diversity.
• Engagement with diverse genomics communities through open-source software, workshops, and integration with public databases to refine the pangenome tools and drive broader adoption.

The award has a performance period from September 2024 through August 2027, with the goal of catalyzing new discoveries by making expansive pangenome resources more accessible and useful for bioinformaticians, geneticists, clinicians, and evolutionary biologists.

Abstract

The human pangenome encompasses global genetic diversity missing from any single reference genome. This project will develop sequence alignment, population genetic, and visualization tools that enable new ways to access and understand the pangenome. These innovations aim to catalyze discoveries by making expansive pangenome resources more intuitive for diverse genomics communities. We will create sequence mapping and alignment methods that allow rapid search and comparison to complement graph-based pangenome approaches. Using succinct data structures, our techniques will enable interpretation of pangenomes while also developing sublinear indexes of the pangenome for sequence search based on exploring all-VS-all genome homology. For population genetics, we will implement analyses for gwas, heritability, selection, and phylogeny directly on a novel encoding of pangenome graphs based on node coverage, which captures allele zygosity across the entire pangenome. This representation naturally captures all variation including everything from snps to structural vari- ants to centromeric haplotypes. To aid exploration, we will build interactive visualizations using GPU-accelerated algorithms to enable real-time interaction with massive graphs which provide a human-scale interface to the overwhelming diversity of data present in pangenomes. Throughout this work, we will engage diverse genomics communities via open source software, hands-on workshops, conferences, and integration with public databases. We will target bioinformaticians, geneticists, clinicians, and evolutionary biologists to refine methodology. Broader pangenome adoption can help overcome reference bias, empower more equitable genomics, and increase UN- derstanding of human genetic diversity to advance health. Our user-focused pangenome tools seek to make these expansive resources tangible and integrated for the genomics community to catalyze new findings.