WinPCA: a package for windowed principal component analysis.

Area of Science:

• Genomics
• Population Genetics
• Bioinformatics

Background:

• Population-scale whole genome sequencing is increasingly common for genomic landscape characterization.
• Traditional methods like FST and dXY provide population-level divergence but lack single-sample resolution.
• Principal Component Analysis (PCA) offers single-sample resolution, aiding in identifying complex genetic structures like inversions and introgression.

Purpose of the Study:

• Introduce WinPCA, a user-friendly Python package for computing, polarizing, and visualizing genetic principal components in genomic windows.
• Provide a tool to analyze genomic variation with single-sample resolution, complementing traditional population genetics statistics.
• Facilitate the identification of genetic structures not aligned with global population structure.

Main Methods:

• WinPCA computes principal components (PCs) in sliding windows along the genome.
• It supports low-coverage whole genome sequencing data by optionally using PCAngsd methods within a genotype likelihood framework.
• The package accepts variant data in VCF or BEAGLE formats.

Main Results:

• WinPCA enables the computation and visualization of genetic principal components across genomic windows.
• The tool provides single-sample resolution for identifying complex genetic variations.
• It generates rich plots for interactive data exploration and presentation.

Conclusions:

• WinPCA is a valuable tool for contemporary genomic studies utilizing genome scans.
• It enhances the analysis of genomic variation by providing single-sample resolution and visualization capabilities.
• The package supports diverse datasets and facilitates the discovery of genetic patterns beyond global population structure.