About me
I am a computational biologist and software engineer, currently holding a Marie Skłodowska-Curie Actions (MSCA) Postdoctoral Fellowship at the University of Edinburgh, hosted by Dr. K. Lohse. My research combines generative AI, mathematical modeling, and software engineering to develop new methods and algorithms for solving complex problems in evolutionary biology.
Before Edinburgh, I was a postdoc at ETH Zürich in the group of Prof. Dr. A. Widmer and at the University of Fribourg with Prof. Dr. D. Wegmann. I received my PhD in Engineering from ITMO University in 2023.
I believe it is essential not only to develop novel methods, but to deliver them as accessible, user-friendly software. I am the lead developer and active maintainer of the award-winning software
GADMA for easy-to-use demographic inference. Moreover, I contribute to community-driven software like
stdpopsim and
demes within the
PopSim consortium. To empower researchers to leverage these tools, I organize and teach the annual
Workshop on Demographic Inference.
Outside of research, I love making science visually accessible. You will often catch me doodling bunnies to explain demographic histories or creating other bizarre illustrations for my talks!
Research
Generative AI for Demographic Identifiability
Developing a generative AI approach to solve the non-identifiability problem in demographic inference. My goal is to deliver a pretrained, ready-to-use AI model. It will discover alternative histories that fit the genetic data equally well — in just seconds. (MSCA Project)
⚙️ Model in active development
Inference of Linked Selection from Time-Series Data
Developed SweepLink, a two-layer Hidden Markov Model framework for the joint inference of linked selection and demography. By explicitly modeling both time and genomic linkage, it achieves state-of-the-art accuracy and detection power for weak selection—a challenging regime for joint inference.
Automated Demographic Inference
Engineered GADMA by developing global optimization algorithms and novel dynamic size-change models for unsupervised demographic inference. This work was supported by a System Biology Fellowship and won a Bronze Humies Award at GECCO and 2nd place at the GHIST 2024 inference tournament.
