Building a Practical Haplotype Graph based Genetic Map of the zea Synthetic Population

Domestication and inbreeding of modern maize has resulted in the loss of much of the genetic diversity found in the wild species of maize, teosinte. Genetic variation in teosinte could provide a useful source of diversity lacking in modern maize cultivars. However, the lack of a high-resolution genetic resource for the study of teosinte has made it difficult to associate that variation with agronomic traits. We are producing a genetic map of the Zea Synthetic maize population, a genetic resource created by randomly mating a 38-accession synthetic breeding population consisting of 11 geographically diverse teosinte (Zea mays ssp. parviglumis) crossed with B73 and the 26 inbred Nested Association Mapping (NAM) founders.

Our approach leverages low coverage GBS data and full genome sequences of the population’s parent genomes using the practical haplotype graph (PHG), a server mounted pangenome built from a SQL database populated by a reference panel of whole genome sequences. To mirror the makeup of our Zea Synthetic population, we filled a reference panel with data from the whole-genome sequences of the 26 inbred NAM founders and 11 teosinte pseudo-genomes. ~4.6 million haplotypes were sampled from individuals and groups in this diverse panel to populate the PHG with individual and consensus reference haplotypes. These haplotypes were paired with low resolution GBS sequences from 2000 doubled haploids from the Zea Synthetic population to accurately impute WGS data for the entire population for use in association mapping.