S7 MagicScan

Alternative

MagicScan performs QTL mapping by a single-QTL linear model via genomic scan. There exist more sophisticated methods using linear mixed models and multi-QTL models. It is recommended to use the statgenMPP package. Here are two key references for statgenMPP: Li et al. 2021 and Li et al. 2022. For more questions, please contact: bart-jan.vanrossum@wur.nl.

MagicScan

The magicscan performs QTL mapping by scanning each marker with the simple linear model that can be specified by keyarg equation. The keyarg equation is a formula expression resulting from @formula. See StatsModels.@formula for the syntax.

For the default equation = nothing, it means lastcolumn ~ 1 where lastcolumn denotes the name of last column in the phenofile. For the default threshold = nothing, the permutation method is used to calculate the threshold for QTL detection.

# code for Julia 
using StatsModels # required for @formula
using MagicScan
cd(@__DIR__)
ancestryfile = outstem*"_magicreconstruct_ancestry.csv.gz"
phenofile = outstem*"_magicsimulate_pheno.csv"
magicscan(ancestryfile,phenofile;
    equation = @formula(phenotype ~ 1 + population),
    outstem     
)

# code for Linux shell. 
# For Window CMD, replace multiline key \ by  ^, and replace comment-key # by ::
julia rabbit_magicscan.jl -g example_magicreconstruct_ancestry.csv.gz \
    -p example_magicsimulate_pheno.csv \
    -e "phenotype ~ 1 + population"  \
    -o example

Output files

Output: peak

using CSV, DataFrames
CSV.read("example_magicscan_peak.csv",DataFrame;comment="##")

using MagicBase
truepheno = MagicBase.readmultitable("example_magicsimulate_truepheno.csv")
truepheno["map_qtl"]

Output: manhattan