Insertion variants missing in the human reference genome are widespread among human populations

Area of Science:

• Genomics
• Human Genetics
• Bioinformatics

Background:

• Structural variants, including insertions, are crucial genomic elements.
• Insertions are historically challenging to detect due to reference bias and complexity.
• Population-level data on non-reference insertions and their functions remain largely unexplored.

Purpose of the Study:

• To develop a computational pipeline for detecting and characterizing non-reference insertion variants.
• To create the first population-level catalog of non-reference insertion variants in humans.
• To investigate the functional implications and evolutionary patterns of these variants.

Main Methods:

• Developed the InserTag pipeline for local de novo assembly and sequence inference of insertion variants.
• Applied InserTag to 2535 individuals from the 1000 Genomes Project.
• Classified variants as ancestral or novel insertions and performed population genetics analyses.

Main Results:

• Identified 1696 non-reference insertion variants, with individuals missing ~0.92 Mbp from the reference genome on average.
• Found that 92% of variants were common, with many being major alleles.
• Observed greater divergence and non-reference sequences in African populations, linked to high-frequency insertions.
• Detected linkage disequilibrium with phenotype-associated SNPs and signals of recent selection.

Conclusions:

• Non-reference insertion variants are a significant component of human genetic diversity.
• The InserTag pipeline offers a robust framework for discovering and genotyping these previously hidden genomic sequences.
• This work lays the foundation for understanding the functional impact of insertion variants in human populations.