Predicting Functional Consequences of Recent Natural Selection in Britain.

Area of Science:

• Human evolutionary genetics
• Paleogenomics
• Molecular evolution

Background:

• Ancient DNA analysis identifies genomic selection signals but struggles to infer phenotypic consequences.
• Noncoding variants significantly contribute to phenotypic effects of selection, making gene expression analysis crucial.

Purpose of the Study:

• To infer phenotypic consequences of natural selection by analyzing changes in predicted gene expression over time.
• To identify genes and genomic loci showing significant expression shifts due to selection in ancient British populations.

Main Methods:

• Utilized Joint-Tissue Imputation (JTI) to predict gene expression from genotype data in ancient and present-day individuals.
• Performed a transcriptome-wide scan on 17,384 protein-coding genes across a 4,500-year time transect.
• Compared results with genome-wide scans estimating per-single nucleotide polymorphism (SNP) selection coefficients.

Main Results:

• Identified 28 genes at seven genomic loci with significant changes in predicted expression (FDR < 0.05) over 4,500 years.
• Highlighted specific genes with expression shifts at previously identified loci and discovered selection on gene expression at two novel loci (SLC44A5 and NUP85).
• Demonstrated that combining classical selection statistics (iHS, SDS) with JTI models can incorporate functional information into present-day data scans.

Conclusions:

• Predicted gene expression analysis using ancient DNA provides a powerful method to explore the phenotypic consequences of natural selection.
• This approach can identify selection acting on gene expression, even at loci not detected by traditional genomic scans.
• Integrating functional data with genomic scans enhances the understanding of both the causes and consequences of human evolution.