A high-throughput screening method for selecting feature SNPs to evaluate breed diversity and infer ancestry

Area of Science:

• Genomics
• Bioinformatics
• Animal Science

Background:

• Whole-genome sequencing (WGS) generates massive single nucleotide polymorphism (SNP) data in livestock.
• Existing methods for analyzing SNP data face overfitting and computational bottlenecks in population genetics studies.
• Efficient selection of informative SNPs is vital for accurate breed diversity assessment and ancestry inference.

Purpose of the Study:

• To develop a novel method, HITSNP, for selecting high-representative SNPs from large genomic datasets.
• To improve the accuracy and computational efficiency of population stratification and ancestry prediction in livestock.
• To introduce a new algorithm for predicting ancestral population structure without pre-determining cluster numbers.

Main Methods:

• HITSNP integrates feature selection techniques with machine learning algorithms.
• The method identifies SNPs that effectively capture information for population diversity and ancestry.
• A novel algorithm is incorporated to predict ancestral population composition and count.

Main Results:

• HITSNP demonstrates superior performance over existing methods in estimation accuracy and computational stability.
• The method effectively selects informative SNPs for breed diversity and ancestry inference.
• HITSNP provides a robust approach for predicting ancestral population characteristics.

Conclusions:

• HITSNP offers a computationally efficient and accurate solution for selecting informative SNPs in livestock genomics.
• The method enhances the study of population structure, breed diversity, and conservation efforts.
• HITSNP facilitates advancements in animal breeding and the protection of animal genetic resources.