Thanks, Rebekah! Happy that our paths crossed as some of the last postdocs from Monty.
Natural selection can provoke substantial biases in predicting complex traits from ancient genomes using polygenic scores, so it’s crucial to consider its impact when predicting complex traits! 7/7
In addition, our simulations showed that both types of selection can hamper polygenic scores accuracy for traits like height and BMI, despite having complete genotype data and perfect effect size estimates. 6/7
This leads to a reduction in ancient polygenic scores accuracy for the first, and a reduction that depends on the allele effects for the later. 5/7
We found that while stabilizing selection accelerates the loss of high-effect alleles, directional selection leads to the loss of alleles that drive phenotypes away from the optimal value. 4/7
We observed that each type of selection creates a different pattern for the alleles that are lost and conserved through time between ancient and modern individuals. 3/7
Using forward in time simulations, we modeled a single additive trait evolving under neutrality, stabilizing selection and directional selection. We examined how each of these evolutionary dynamics affects ancient polygenic scores accuracy. 2/7
Thread by Valeria Añorve-Garibay reproduced here: We used a simulation-based approach to explore how natural selection acting on complex traits impacts the predictive accuracy of polygenic scores when applied to ancient samples. 1/7
Our preprint analyzing the impact of stabilizing and directional selection on the prediction of complex traits in ancient samples is out. Work spearheaded by Valeria Añorve-Garibay in collaboration with Mashaal Sohail and Emilia Huerta-Sánchez. Thread below: www.biorxiv.org/content/10.1...
bioRxiv - the preprint server for biology, operated by Cold Spring Harbor Laboratory, a research and educational institution