HIBLUP currently only accept the genotype in PLINK binary format, e.g.
demo.bed, please see more details at PLINK user manual. Users can convert any other format of genotype (e.g. VCF, HapMap, PED/MAP) to binary format by PLINK2 and TASSEL:
# HapMap to VCF by TASSEL ./run_pipeline.pl -SortGenotypeFilePlugin -inputFile demo.hmp.txt -outputFile demo.sort.hmp.txt -fileType Hapmap ./run_pipeline.pl -fork1 -h demo.sort.hmp.txt -export-exportType VCF -runfork1 # VCF to Binary ./plink2 --vcf demo.vcf --make-bed --out demo # PED/MAP to Binary ./plink2 --ped demo.ped --map demo.map --make-bed --out demo
Those binary files should be assigned to flag
(1) HIBLUP only supports to load one binary file, does not support to load more than one binary files (e.g. whole genome is stored separately by chromosome), please merge them by PLINK prior to using HIBLUP, a rough guidance can be found here.
(2) HIBLUP will code the genotype A1A1 as 2, A1A2 as 1, and A2A2 as 0, where A1 is the first allele of each marker in
*.bim file, therefore the estimated effect size is on A1 allele, users should pay attention to it when a process involves marker effect.
HIBLUP has no function of imputation, any missing genotype in binary file will be treated as heterozygote, that means missings will be forced to be coded as 1 in analysis, we suggest users to implement imputation by other software prior to using HIBLUP if missings exist in genotype.