editor_options
markdown wrap 72

hbecher_ASR2kmerGwas

Code to convert ASR simulations to k-mer datasets

Workflow

1. Run Rscript 10ASR2FASTA.R.
2. This will produce:

• One reference genome FASTA file fatsa/ReferenceGenome.fa. All nucleotide positions in other files are given relative to this file.
• 3000 FASTA files of simulated individuals (multi-chromsome, diploid genomes)
  • For more files, up to 33000, comment out line 126 (haploFiles <- dir(pattern="haplotypes*PYT")) of file 10ARS2FASTA.R)
• AllGvData.txt (tab-separated, 1st column matches the FASTA file names)
• AllPhenoData.txt (tab-separated)
• AllVarLociAddPart.txt (tab-separated) with one column for each trait, showing the additive effect that each variant site has on each trait. This file can also be used to match variant IDs (1st column) to nucleotide positions in the reference (cols 2 and 3).
• Four per-trait QTL files. These also list the pairwise epistatic interaction effect for each trait.

Traits

Four traits ares simulated:

• Trait1: purely additive, gaussian
• Trait2: additive + epistatic, gaussian
• Trait3: purely additive, gamma
• Trait4: additive + epistatic, gamma

The additive contributions of each variant site (position with respect to the reference genome), are written to AllVarLociAddPart.txt. For the epistatic traits, the pairwise interaction effects are written to the per-trait QTL files, QTL[...]txt.

For k-mer GWAS

Use one of the gv.Trait... columns from file AllGvData.txt as the phenotypes. The do not have any environmental randomness. For more environmental noise, you can use pheno.Trait... from file AllPhenoData.txt instead.

Requirements

• R: AlphaSimR, dplyr, here

Code sources

This uses code from https://github.com/Rabab53/ggenerator.