Design parameters for genome editing in humans Event as iCalendar

(Biological Science, Computer Science)

31 August 2017

11am - 12pm

Venue: Building 303S, Room 561

Location: City Campus

Host: Center for Computational Evolution

Cost: Free

Contact email:



Presenter: Alexei Drummond, Professor of Computational Biology, CCE Director


It has been claimed by some that CRISPR-Cas9 genome editing produces results indistinguishable from natural and artificial breeding. Furthermore, recent success in using CRISPR-Cas9 to edit the genomes of human embryos has laid the foundation for the first gene-modified people and led some researchers to claim that "gene editing could be used as a preventive tool to lower a newborn’s lifetime risk of a wide range of diseases, including cancer". Here I investigate both of these claims from a theoretical perspective. 

First I will review the literature on genetic load in humans. I will show some theoretical results under models of genetic load for parameter values relevant to the human population. In doing so I will demonstrate a simple and reasonable scenario in which genome editing will produce genomes that could never be produced by normal population genetic processes, despite utilising only existing genetic variation. 

I will also describe genetic models of pleiotropy and epistasis - processes fundamental to understanding the genetic basis of complex diseases. These models suggest that genome-editing may not be able to simultaneously lower the risk of all complex diseases, thus converting the use of genome editing as a preventive tool into a difficult multi-objective optimisation problem. 

I will finish by describing research my group is pursuing to understand the potential population-level consequences of the application of genome editing to improving human health.