Effective models for genetic data analysis

Martina Favero received her doctoral degree in mathematical statistics from KTH Royal Institute of Technology in 2021. Thanks to a grant from Knut and Alice Wallenberg Foundation, she will hold a postdoctoral position with Professor Paul Jenkins at the University of Warwick, Coventry, United Kingdom.

What do DNA sequences tell us about human history? When were horses domesticated? How quickly does the coronavirus mutate? These are examples of questions studied in population genetics, a field that was created by merging two seemingly incompatible theories from the mid-nineteenth century: Darwin’s theory of evolution, which is based on natural selection and chance, and Mendel’s theory of inheritance, describing how qualities are passed down from parents to children.

The mathematical models currently used in population genetics originate from the early twentieth century. These models brought both theories together, to explain how a population’s genetic composition evolves over time due to the action of natural selection, mutations, mating, migration and other factors. Random variations play a great role in the evolution of a population, so the theory of probability is a valuable tool in determining how genetic traits vary in a population, for example.

Although there are several methods suitable for analyzing genetic data, the vast amounts of data nowadays available due to advances in DNA sequencing technology create new challenges. Too much data is a problem for the classical statistical methods, because often the existing algorithms either result in calculations being practically unfeasible, or provide inaccurate estimates.
Therefore, the standard techniques need improvement and new models must be developed. The purpose of Martina Favero’s project is to create new mathematical methods for more effective and more reliable calculations in population genetics.