Population clustering based on copy number variations detected from next generation sequencing data.

Area of Science:

• Genomics
• Bioinformatics
• Population Genetics

Background:

• Copy number variations (CNVs) are valuable biomarkers detectable with high resolution by next-generation sequencing (NGS).
• Extracting informative features from CNVs for genomic studies, particularly population clustering, remains a significant challenge.

Purpose of the Study:

• To develop a novel method for population clustering utilizing CNVs derived from NGS data.
• To address the challenge of feature extraction from CNVs for effective genomic analysis.

Main Methods:

• Extraction of CNVs from each sample to create a feature matrix.
• Application of Non-negative Matrix Factorization (NMF) to decompose the feature matrix into source and weight matrices.
• Identification of common CNVs (source matrix) and their abundance (weight matrix) within samples for differentiation.

Main Results:

• The proposed NMF-based method successfully clustered simulated data, accurately recovering common CNVs.
• Analysis of real data from the 1000 Genomes Project demonstrated effective clustering of individuals from different ethnic groups.
• Validation on a large-scale, whole-genome dataset confirmed the method's applicability and robustness.

Conclusions:

• Non-negative matrix factorization of CNVs provides a powerful approach for population clustering.
• The method effectively differentiates ethnic groups by leveraging shared CNV patterns.
• This technique shows significant potential for large-scale population genetic studies using NGS data.