Detection and classification of positive selection in human populations

Detecting positive selection in genomic regions is a recurrent topic in human population genetics studies. Over the years, many positive selection tests have been implemented to highlight specific genomic patterns left by a selective event when compared to neutral expectations. However, there is lit...

Descripción completa

Detalles Bibliográficos
Autor: Pybus Oliveras, Marc
Tipo de recurso: tesis doctoral
Estado:Versión publicada
Fecha de publicación:2015
País:España
Institución:CBUC, CESCA
Repositorio:TDR. Tesis Doctorales en Red
OAI Identifier:oai:www.tdx.cat:10803/384314
Acceso en línea:http://hdl.handle.net/10803/384314
Access Level:acceso abierto
Palabra clave:Positive selection
Machine-learnong
Human population genetics
Simulations
Selective sweep
Selecció positiva
Aprenentatge supervisat
Genètica de poblacions humanes
Simulacions
575
Descripción
Sumario:Detecting positive selection in genomic regions is a recurrent topic in human population genetics studies. Over the years, many positive selection tests have been implemented to highlight specific genomic patterns left by a selective event when compared to neutral expectations. However, there is little consistency among the regions detected in several genome-wide scans using different tests and/or populations: population-specific demographic dynamics, local genomic features or different types of selection acting along the genome at different times and selective coefficients might explain such discrepancies. The present doctoral thesis is focused in the study of this problem and the development of a innovative solution: a machine-learning classification framework that exploits the combined ability of some selection tests to uncover the different features expected under the hard sweep model, such as sweep completeness and age of onset. The method was calibrated and applied to three reference populations from The 1000 Genome Project to generate a genome-wide classification map of hard selective sweeps. This study improves the way a selective sweep is detected by overcoming the classical selection vs. no-selection classification strategy, and offers an explanation to the lack of consistency observed among selection tests when applied to real data.