GWAS meta-analysis of aPTT as a way to understand coagulation.

This repository contains supplementary material (scripts, figures, and tables) used during the realization of my Final Master Project, titled: GWAS Meta-analysis of aPTT as a Way to Understand Coagulation.

Workflow Diagram of the procedures conducted in this project:

Contents

• Scripts: Code used for data processing, statistical analysis, and visualization.
• Figures: Graphs and plots generated during the analysis.
• Tables: Data tables and summary statistics.

Objective

The main goal of this project is to identify genetic variants or SNPs associated with aPTT through a comprehensive meta-analysis of genome-wide association studies (GWAS), using GWAS data from multiple cohorts provided by The Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) consortium.

Usage

The scripts and data provided here are intended for reproducibility and further analysis. Feel free to explore and adapt the material for related studies or projects.

1. GWAS data preprocessing

The code used during the data harmonisation of the provided cohorts is included in the R folder with the name GWAS_data_processing.R In this procedure, we made an inverse normal transformation in order to have the same beta and standard error scale within every dataset.

2. Quality Control performed with EasyQC and Meta-analysis performed with METAL software

After harmonisation, we performed quality control using EasyQC package from R. Model scripts with the corresponding quality standards are located in the tools folder with the name EasyQC_script.ecf. Then, we performed the meta-analysis using METAL software. Model scripts used for this tool is located in the tools folder with the name MetalScript.txt

3. Top tables

To process meta-analysis dataframes and extract the top SNPs, use the script located in the R folder: script_toptable_metaanalysis.R

TWAS

For TWAS analysis, follow these steps:

• Data Harmonization: Harmonize data from meta-analysis using the script located in the BASH folder: harmonization_script.sh
• Data Imputation: Impute data with the following script, also found in the BASH folder: imputation_script.sh
• Harmonized and Imputed Data Merging: Merge harmonized data with imputed data using the script: run_imputation.sh

Prediction and Meta-analysis (PrediXcan and MetaXcan softwares):

• prediXcan Software: Use the script spredixcan_script.R located in the R folder.
• MetaXcan Software: Use the script smultixcan_script.sh located in the BASH folder.

Visualization

Visualization scripts are available in the plots folder:

• PrediXcan z-score plot: zscoreplot_predixcan.R
• Top SNPs Manhattan plot: topsnps_manhattan_plot.R