The draft genome of Ditylenchus dipsaci.

Ditylenchus dipsaci is a devastating pest to many crops worldwide. We present the first genome sequence for this species, produced with PacBio sequencing and assembled with CANU. Ditylenchus dipsaci is a devastating pest to many crops worldwide. We present the first genome sequence for this species, produced with PacBio sequencing and assembled with CANU.

The stem and bulb nematode, Ditylenchus dipsaci, is a plant-parasitic nematode affecting over 500 plant species with major destruction occurring in onion, garlic, beet, strawberry, broad bean, potato, alfalfa, and oat (Sturhan and Brzeski, 1991). It is a quarantine pest in Europe (EPPO, 2004) and well known in garlic production where it can cause losses of up to 90% (Abawi and Moktan, 2010). The species exhibits a variety of host preferences for which biological groups were proposed (Sturhan and Brzeski, 1991). In this study, we report the first genome assembly of D. dipsaci, a resource that will improve our understanding of this important parasite.

A population of D. dipsaci (E-105) was obtained from the Canadian National Collection of Insects, Arachnids, and Nematodes and originally isolated from garlic in Northern Ontario (Qiao et al., 2013). The nematodes were multiplied in vitro on pea sprout inoculated with only 15 mixed-stages individuals, as described in Poirier et al. (2019), to reduce heterozygosity. DNA was extracted from ~7,500 mixed-stage D. dipsaci, using the Qiagen DNeasy Blood & Tissue Kit (Valencia, CA). MUGQIC (Montréal, Canada) performed the library preparation (sheared large insert) and sequencing on a PacBio RSII instrument using 12 SMRT cells yielding 941,946 long reads at a mean size of 10,207 bp for an estimated coverage of 42×. Two assemblies were generated using CANU v1.7 (Koren et al., 2017), one with default parameters and one optimized for high heterogeneity (corMhapSensitvity = high, corMinCoverage = 0, and correctedErrorRate = 0.105). These pre-assemblies were merged using Quickmerge (Chakraborty et al., 2016) and decontaminated using Blobtools v1.0 (Laetsch and Blaxter, 2017). Before scaffolding, the assembly contained 3,559 contigs with a N50 of 152 kb. Contigs were then scaffolded using MeDuSa v1.6 (Bosi et al., 2015) with the genomes of Caenorhabditis elegans (GCA_000002985.3; The C. elegans Sequencing Consortium, 1998), D. destructor (GCA_001579705.1; Zheng et al., 2016), Globodera pallida (GCA_000724045.1; Cotton et al., 2014), G. rostochiensis (GCA_900079975.1; Eves-van den Akker et al., 2016), and Meloidogyne incognita (GCA_900182535.1; Abad et al., 2008). Scaffolding was further improved using Rascaf (Song et al., 2016) and 13,555,476 PE RNA-seq reads (Phred score > 20), obtained from mixed-stage D. dipsaci (Illumina HiSeq 2 × 100 bp, TruSeq Library Prep Kit). Then, Purge Haplotigs (Roach et al., 2018) resolved highly polymorphic regions and reduced artifactual duplication.

The D. dipsaci genome assembly is comprised of 1,394 scaffolds, a largest scaffold of ~3.6 Mb, 22 scaffolds >1 Mb, and a N50 of 287 kb. The assembly is 227,234,012 bp (227.2 Mb), double the size of the D. destructor assembly (111.1 Mb). We used flow cytometry on D. dipsaci nuclei to investigate this discrepancy, revealing a haploid genome size of 160 to 170 Mb. Thus, obtaining 227 Mb with heterozygous material is not surprising. The GC content of the D. dipsaci assembly was 37.5% which is comparable to D. destructor (36.8%), C. elegans (35.4%), G. rostochiensis (38.1%), and G. pallida (36.7%). A completeness assessment with BUSCO v 3.0.2 (Simão et al., 2015) using the Nematoda_odb9 database revealed 537 complete sequences (506 single-copy, 31 duplicated), 133 fragmented, and 312 missing on a total of 982 BUSCO genes searched in D. dipsaci. The number of complete genes was higher than G. pallida (431) but lower than D. destructor (749). Gene prediction was performed with Braker 2.1.0 (Hoff et al., 2015) on a genome masked with Repeatmodeler/Repeatmasker 4.0.7 (Chen, 2004), leaving simple repeats unmasked. Transcript information was relayed to Braker via HISAT2 2.1.0 alignments of D. dipsaci RNA-seq and exonerate 2.4.0 aligned Ditylenchus destructor proteins to produce a gene annotation of 26,428 putative genes. On average, these genes contained seven exons (max = 114) with a mean length of 120 bp (from 2 to 9,194 bp) and six introns (max = 113) with a mean length of 515 bp (from 36 to 27,260 bp).

GenBank accession numbers: The raw sequences and genome assembly files are available under the NCBI BioProject PRJNA498219.