diploid_potato_qtl_mapping

These are the QTL mapping scripts, genotypic data, and phenotypic data that accompany Marand et al. Residual heterozygosity and epistatic interactions underlie the complex genetic architecture of yield in diploid potato. 20XX. doi:xxxxxx

QTL mapping scripts

QTL mapping in these scripts is performed by fitting multiple linear regression at each recombination bin independently. The model contains terms for each parental allele (US-W4 and M6), as well as an interaction term. In bins where one parent is not segregating, that parent and the interaction term are both excluded from the model.

• 1_run_dual_interaction_mapping.R: This is the primary script that will run QTL mapping for all traits, write the results in text form, and produce plots of significance profiles.
• map_dual_interactions.R: contains functions to construct the QTL mapping model (make_model()) and to apply the model across all recombination bins for a single phenotype (dual_mapping()).
• run_perm.R: contains a single function, runPerm(), that shuffles the phenotypes and runs QTL mapping.

Genotypic and phenotypic data

• W4M6_haplotype_bins_split.bed: Haplotype designations for each individual at each recombination bin. Columns 1-3 contain chromosome, haplotype bin start coordinate, and haplotype bin end coordinate. Subsequent pairs of columns contain binary scores for W4 and M6 alleles, respectively. For example, chr01 0 873946 1 0 corresponds to a bin on chromosome 1 spanning the physical coordinates from 0 to 873946. The individual has W4 haplotype 1 and M6 haplotype 0. The maternal and paternal parents are W4 (highly heterozygous) and M6 (seven-generation inbred), respectively.
• yield_phenological_traits_allGenos.csv: contains BLUPs for yield, tuber weight, and tuber number; intercept and slope terms from longitudinal random regression modelling of plant height; and the binary response for growth habit (labelled as 'Height_Group').