Construction of a High-Density SNP-Based Genetic Linkage Map of the Ipomoea trifida (H. B. K.) G. Don. by Scaffolding a de novo Assembly

Despite being a critical worldwide food resource, the sweetpotato (Ipomoea batatas (L.) Lam.) remains a less well studied crop due to its genetic complexity (2n=6x=90). As a result the availability of sweetpotato genomic resources is highly limited. Here we propose a novel method for constructing genetic linkage maps and scaffolding de novo assemblies using Genotyping-by-Sequencing (GBS) data from a large full-sib family. We have used this approach to construct a high-density SNP-based genetic linkage map consisting of 7,799 SNPs from 15 linkage groups by scaffolding a de novo assembly of Ipomoea trifida (H. B. K.) G. Don., the most likely ancestor of the sweetpotato but of much less genetic complexity (2n=2x=30). We started from inferring transmissions of parental haplotypes to each progeny, from which we were able to calculate the recombination fraction (RF) between every pair of contigs. We then iteratively clustered contigs into linkage groups, and finally resolved the contig ordering by minimizing the RF sum along each linkage map. Our method also uses the shared haplotype structure in the family together with the allelic depth information to call high quality SNP genotypes. Application of our approach to simulated datasets indicates that we are able to accurately reconstruct the correct clustering and ordering of contigs. Moreover, as is capable of solving high ploidy genomes, we anticipate this method may help in constructing high quality genetic linkage maps for more plants.