Genome Improvement and Genetic Map Construction for Aethionema arabicum, the First Divergent Branch in the Brassicaceae Family

Abstract

The genus Aethionema is a sister-group to the core-group of the Brassicaceae family that includes Arabidopsis thaliana and the Brassica crops. Thus, Aethionema is phylogenetically well-placed for the investigation and understanding of genome and trait evolution across the family. We aimed to improve the quality of the reference genome draft version of the annual species Aethionema arabicum Second, we constructed the first Ae. arabicum genetic map. The improved reference genome and genetic map enabled the development of each other. We started with the initially published genome (version 2.5). PacBio and MinION sequencing together with genetic map v2.5 were incorporated to produce the new reference genome v3.0. The improved genome contains 203 MB of sequence, with approximately 94% of the assembly made up of called (non-gap) bases, assembled into 2,883 scaffolds (with only 6% of the genome made up of non-called bases (Ns)). The N₅₀ (10.3 MB) represents an 80-fold increase over the initial genome release. We generated a Recombinant Inbred Line (RIL) population that was derived from two ecotypes: Cyprus and Turkey (the reference genotype. Using a Genotyping by Sequencing (GBS) approach, we generated a high-density genetic map with 749 (v2.5) and then 632 SNPs (v3.0) was generated. The genetic map and reference genome were integrated, thus greatly improving the scaffolding of the reference genome into 11 linkage groups. We show that long-read sequencing data and genetics are complementary, resulting in an improved genome assembly in Ae. arabicum They will facilitate comparative genetic mapping work for the Brassicaceae family and are also valuable resources to investigate wide range of life history traits in Aethionema.

Keywords: Aethionema arabicum; Brassicaceae; Genotyping by Sequencing; MinION; PacBio; genetic map; genome improvement.

Figure 1

Overview of the analyses performed in this study. In filled boxes are data sets, approaches and companying tools are in open boxes.

Figure 2

Problem arising from applying PBjelly2 on vAM. Scaffold borders are visualized in blue and extensions of scaffolds introduced by PBjelly2 are shown in brown. Assuming the true order of the scaffolds is shown on top of the figure, but scaffold X and scaffold Z were already combined in the vAM assembly (second bar from top) this could lead to a partial filling of the N-stretch and maybe an extension of scaffold Y. However, PBjelly2 would not be able to place scaffold Y between the two other scaffolds (middle bar). If the scaffolds were thus split again (second bar from bottom), it is possible that the connections are made correctly applying PBjelly2 on the split version (bottom bar). This only visualizes a theoretical case, in this work it appeared every time that scaffold X and Y were connected by PBjelly2 and scaffold Z had to be reconnected afterward.

Figure 3

Aethionema arabicum genetic map v2.5. Genetic map version 2.5 consists of eleven linkage groups. On each linkage group, genetic distance in cM is present on the left and SNP markers on the right.

Figure 4

Aethionema arabicum genetic map v3.0. Genetic map version 3.0 consists of eleven linkage groups. On each linkage group, genetic distance in cM is present on the left and SNP markers on the right.

Figure 5

The alignment of genetic map v2.5, v3.0 and physical map. The alignment of the genetic map v2.5 and v3.0 were based on relative SNPs. The left ruler indicates genetic distance in cM and the right indicates physical distance in bp according to genome v3.0.