Sebastian graduated with a MSc in Probability at CIMAT in Guanajuato Mexico in 2018. During his studies he focused on Markov Chains and the use of Monte Carlo methods for Bayesian inference. After this he worked for 3 years as a data scientist in industry, using a variety of computational and statistical models to optimize sales and transportation routes. This working period, together with his maths background, motivated his desire for learning about the underlying mathematical theory behind the various computer algorithms, as well as their application in other disciplines. In his spare time, he enjoys reading fiction, watching movies, and playing football and racket sports.

Research project title:

Monte Carlo Methods for Bayesian Inference of Ancestral Recombination Graphs

Supervisor(s):

Kari Heine

Project description:

Knowing the genealogical history of a given species helps us understand its evolution better. The genealogical history can be formally encoded into a graph known as the ancestral recombination graph (ARG). However, the ARG is never available in practice, and inferring the ARG from sequenced DNA data is a challenging problem since the computational time grows hyper-exponentially. The research objective is to study existing algorithms such as the ARGweaver and ARGinfer, as well as the conditional sequential Monte Carlo method, to make improvements for the Arbores algorithm.