A pangenome-graph approach for mapping and imputing barley sequences

Barley (Hordeum vulgare) is a crucial cereal crop capable of thriving in various environments due to its remarkable adaptability. However, its extensive, repetitive genome poses challenges for representing pangenomes. Starting from the reference genome MorexV3, researchers have developed a barley graph called Pan20, which encompasses the genetic diversity of barley landraces and cultivars found in the public pangenome V1.

A novel method for mapping arbitrary sequences to the barley graph is proposed. This approach combines a greedy strategy involving GMAP alignment, followed by an intersection with a Practical Haplotype Graph (PHG). This technique allows for the detection of presence-absence variation and provides a consistent physical coordinate system based on the MorexV3 reference genome across different barley genotypes. This enables comparative analysis and visualization of barley sequences.

For imputing genomic data, the PHG method employs k-mer pseudo-alignment against the barley graph. Benchmarks demonstrate that Pan20 can accurately align barley genomic and transcriptomic sequences, even those not represented in the Morex reference genome. The results show that a third of long genomic sequences map onto non-reference genomes.

Additionally, experiments with Genotyping by Sequencing and low-pass sequencing data reveal that FASTQ files can be efficiently mapped and imputed against the graph while preserving local haplotype context.

This flexible and scalable graph framework empowers barley researchers to investigate genetic diversity beyond a single reference genome. It facilitates the analysis of diversity panels at the haplotype level, going beyond the limitations of SNP analysis. Documentation and a Docker container are accessible at https://github.com/eead-csic-compbio/barleygraph. The graph sequence mapping utility has also been integrated into a web application available at https://barleymap.eead.csic.es.