Cluster scalable pangenome graph construction with nf-core/pangenome

Pangenome graphs can represent variation between multiple genomes. The PanGenome Graph Builder (PGGB) (Garrison, Guarracino et al., bioRxiv 2023) is a method for constructing unbiased pangenome graphs that fully represent genomic variation without excluding any sequences. PGGB uses all-to-all genome alignments and learned graph embeddings to build and iteratively refine a model in which we can identify variation, measure conservation, infer phylogenetic relationships, and detect recombination events (Guarracino et al., Nature 2023). However, PGGB is deployed as a bash script, limiting its scalability to only one cluster node.

Here we present nf-core/pangenome, a cluster scalable implementation of PGGB. nf-core/pangenome can distribute genome alignments across multiple nodes of a cluster scaling to up to thousands of input sequences. We show its automated sequence partitioning feature by building a Lodderomyces elongisporus pangenome graph, aiming to assign each chromosome into one graph component. A detailed MultiQC report summarizes graph statistics and shows diagnostic graph visualizations in 1 and 2 dimensions. nf-core/pangenome is written with Nextflow’s DSL2 and incorporates nf-core modules, ensuring modularity while adhering to nf-core community standards.