Pyro: A Comprehensive Pipeline for Eukaryotic Genome Assembly

Abstract

1The assembly of reference-quality, chromosome-level genomes for both model and novel eukaryotic organisms is an increasingly achievable task for single research teams. However, the broad variety of sequencing technologies, assembly algorithms, and post-assembly processing tools currently available means that there is no clear consensus on a best-practice computational protocol for eukaryotic de novo genome assembly. An ever-increasing field of algorithms and packages with unique parameters, setup requirements, and environments makes it difficult for groups to pick up and test new tools, despite potential benefits. Here, we present a comprehensive Snakemake-based pipeline for eukaryotic genome assembly, Pyro, to further assist future de novo assembly and benchmarking projects. Pyro combines 20 assembly and eight polishing packages, comprising 30 different assembly approaches and up to 48 different polishing approaches in combination. These are available across Illumina short-read, Nanopore and PacBio CLR long-read technologies in one container, complete with data preparation, quality metric calculation and result reporting. We demonstrate Pyro effectiveness by running Pyro on publicly available Illumina, Nanopore and PacBio CLR read sets for Arabidopsis thaliana, producing 12 candidate assembly options with minimal initial input or configuration, each with extremely high contiguity and completeness. Pyro is highly customizable to expert needs, while also providing an accessible suggested set of tools for more casual users based on simple inputs. Pyro is available as a Singularity container suitable for execution on any Unix-compatible OS, and is freely available on GitHub (https://github.com/genomeassembler/pyro). This pipeline provides a one-stop solution for a variety of de novo eukaryotic genome assembly needs, and will also assist in the assessment of new tools as a convenient benchmark-generating platform.