N. Cooke 1,*, D. Bradley 2, S. Nakagome 1
2Genetics, Trinity College Dublin, Dublin, Ireland
Topic: Genome-wide methods for detecting selection
The impact of natural selection on beneficial alleles can be observed in modern human genetic variation; however deciphering the origins of these alleles is made difficult by the vast complexity of human history. Here we describe a new statistical framework of Approximate Bayesian Computation (ABC) that can detect which ancestral lineage an allele undergoing selection first appeared. To demonstrate this framework we assume a model based on the history of modern Japanese, in which the present-day population derives ancestry in unequal proportions from two groups, the Jomon and the Yayoi, who admixed with each other 3,000 years ago, having previously diverged 30,000 years ago. By comparing the present-day genetic variation observed at a selected allele to data simulated within our model, we test if our approach can accurately predict which ancestral population it first emerged in. In this presentation, we demonstrate the power of our framework using a simulation study and the application of our approach to modern Japanese genomic data.
Disclosure of Interest: None Declared