This project extracts genomic variants from VCF files to compute Polygenic Risk Scores (PRS) for African populations. Using bcftools and PLINK2, filtered variants are matched to a coronary artery disease PGS (PGS004696). The workflow includes sample filtering, PRS computation, and visualizing PRS distributions across populations.
## Project 7: VCF Variant Extraction and PRS Analysis
Project Overview
This project focuses on analyzing genomic data to extract variants and compute Polygenic Risk Scores (PRS) using bcftools and PLINK2. The analysis highlights African population samples from chromosome-level data files, processes relevant variants, and visualizes PRS distributions.
Table of Contents
Commands Overview
Workflow Summary
Tools and Software
Data Files
PRS Computation
Visualization Outputs
1. Commands Overview
Key Tasks
Extract First 10 Variant IDs:
Filter chromosome 15 data to display the first 10 variant IDs.
Count Unique Samples:
Query and count distinct sample IDs from the VCF file.
Retrieve Total Number of Variants:
Compute the total number of records (variants) for chromosome 12.
2. Workflow Summary
The following steps summarize the workflow:
Variant Extraction:
Extract African-specific samples for chromosomes 1–22.
Save filtered VCF files.
Unique Sample Identification:
Process all VCF files to identify unique African samples.
PRS Analysis:
Download the PGS score table.
Extract rsID for relevant variants.
Filter overlapping variants between the PGS table and VCF data.
PRS Calculation:
Convert filtered VCF files to PLINK2 format.
Calculate PRS for all chromosomes.
Population Merging:
Merge PRS results with population metadata.
Filter data for African populations.
Visualization:
Generate density plots, boxplots, and violin plots for scaled PRS scores.
3. Tools and Software
bcftools: For VCF file filtering and querying.
PLINK2: For PRS calculations and genetic data processing.
Python Libraries:
pandas for data manipulation.
matplotlib and seaborn for data visualization.
subprocess for running system commands.
4. Data Files
Input Files:
VCF Files: Chromosome-level VCF files located at /storage/ice-shared/biol6150/Data/1000Genomes/.
PGS Table: PGS004696 Harmonized data for GRCh38 genome build.
Population File: kgp3_id_pop.tsv containing sample population metadata.
Output Files …