modelling_hpp

Installing PLINK 1.9

PLINK 1.9 is a powerful and efficient tool for whole-genome association studies. This guide provides step-by-step instructions for downloading and installing PLINK 1.9 on a Linux system.

---

1. Download PLINK 1.9

Download Plink 1.9

unzip plink_linux_*.zip
cp plink_install/plink ~/bin/
chomd 755 ~/bin/plink
export PATH=~/bin:$PATH
source ~/.bashrc
plink --version

Running the Script

An example scripts can be found in gwas_v1.sh , which streamlines genotype data quality control, GWAS analysis, and significant SNP extraction.

   bash run_gwas.sh

---

Input Files

1. merge_list.txt

• A text file listing the PLINK binary files (.bed, .bim, .fam) to be merged.

  • .bed: Binary genotype file.
  • .bim: SNP information file.
  • .fam: Sample information file.

• Format: Each line should specify the path to the binary files without the file extensions. Example:

  dataset1
  dataset2
  dataset3
  

2. phenotype.txt

• A tab-delimited file containing phenotype data.
• Format:
  • The first two columns must match the FID (Family ID) and IID (Individual ID) in the .fam file.
  • Subsequent columns represent phenotypes. Each column header should indicate the phenotype name.
• Example:

  FID IID Height Disease
  1   1   165.2    1
  1   2   170.8    0
  2   1   160.5    1
  

• Continuous Phenotypes: Values like height or BMI.
• Binary Phenotypes: Encoded as 1 (control) and 2 (case).

---

Key Features

1. File Merging:

   • Combines genotype files based on a merge_list.txt.

2. Quality Control:

   • Filters genotype data based on missingness (GENO), minor allele frequency (MAF), and Hardy-Weinberg equilibrium (HWE).
   • GENO: Missingness rate (default: 0.05).
   • MAF: Minor allele frequency threshold (default: 0.05).
   • HWE: Hardy-Weinberg equilibrium threshold (default: 1e-6).

3. GWAS Analysis:

   • For each phenotype listed in the phenotype.txt file:
   • If the phenotype is continuous, linear regression is performed.
   • If the phenotype is binary, the pipeline ensures values are encoded as 1 (control) and 2 (case), converting from 0/1 if necessary. Logistic regression is then performed.

4. Significant SNP Filtering:

   • Extracts SNPs with a p-value < 5e-8 and saved in the significant_snps directory.

Output Files

• Quality-Controlled Genotype Data: qc_results/qc_data.*
• GWAS Results: gwas_results/gwas_<phenotype_name>.*
• Significant SNPs: significant_snps/<phenotype_name>_significant_snps.txt