The Model-Based Evolutionary Genomics Unit works at the interface of computational and evolutionary biology. Currently, our two main focus areas are:
i) reconstructing the Tree of Life, including the history of early life forms by continuing to develop and apply state-of-art probabilistic models of molecular evolution and using machine learning to model co-evolutionary dependencies across biological scales to reconstruct ancient phenotypes and environments;
ii) understanding somatic evolution in hierarchically organised tissues and across the Tree of Life, both from a theoretical standpoint (e.g., To what extent has tissue organisation evolved to minimise somatic evolution humans? Why do both plants and animals have stem cells?) and from a more data orientated perspective (e.g., What can emerging data on genetic variation in healthy tissues tell us about tissue organisation, the emergence of tumours and ageing?).
Recent papers of ours concerning these topics include:
A Geological Timescale for Bacterial Evolution and Oxygen Adaptation
Davin, Woodcroft, ..., Szánthó, ..., Fischer, Donoghue, Spang, Hugenholtz*, Williams*, Szöllősi* Science 2025 https://www.science.org/doi/10.1126/science.adp1853 (free author link)
Assuming that aerobic bacteria likely emerged after the Great Oxidation Event, our recent study not only better resolves the age of major bacterial groups, but also finds that oxygen utilisation predated its atmospheric rise by approximately 900 million years, potentially facilitating the evolution of oxygenic photosynthesis.
In collaboration with Phil Donoghue and Tom Williams, with Chris Kay's lead we date gene-duplication events across eukaryogenesis to infer a long protracted assembly of eukaryotic cellular complexity, with many host-cell innovations predating mitochondrial endosymbiosis.
Gene copy-number features generalize better than SNPs for antimicrobial resistance prediction in Staphylococcus aureus
Fistarol, Gervasio, Szöllősi npj Antimicrobials and Resistance 2025 https://www.nature.com/articles/s44259-025-00172-6
In Bruna's first paper we show that pan-genome gene copy-number (including absence) yields more accurate and substantially better out-of-lineage generalization for AMR prediction than core-genome SNP-based features, across 4,255 isolates and six antibiotics.
A timetree of Fungi dated with fossils and horizontal gene transfers
Szánthó, Merényi, Donoghue, Gabaldón, Nagy, Szöllősi, Ocaña-Pallarès Nature Ecology & Evolution 2025 https://www.nature.com/articles/s41559-025-02851-z
Lead by Lénárd and Eduardo Ocaña-Pallarès we reconstruct and date a large-scale fungal phylogeny using fossil calibrations plus relative constraints from fungi-to-fungi HGT, yielding an updated timescale for crown Fungi and early fungal–plant-associated ecological interactions.
Phylogenetic Reconciliation: Making the Most of Genomes to Understand Microbial Ecology and Evolution
Williams, Davin, Szánthó, Stamatakis, Wahl, Woodcroft, ..., Spang, Hugenholtz, Szöllősi The ISME Journal 2024 https://academic.oup.com/ismej/article/18/1/wrae129/7713227
In this review we explore how phylogenetic reconciliation can serve as a powerful tool to study genome evolution, offering insights into ancestral gene content and metabolic pathways, thereby enhancing our understanding of microbial ecology and evolution.
Divergent Genomic Trajectories Predate the Origin of Animals and Fungi
Ocaña-Pallarès, Williams, …, Bapteste, Tikhonenkov, Keeling, Szöllősi, Ruiz-Trillo Nature 2022 https://www.nature.com/articles/s41586-022-05110-4
Our paper reveals that the distinct genetic paths leading to animals and fungi were established before the emergence of multicellularity.
Trade-off Between Reducing Mutational Accumulation and Increasing Commitment to Differentiation Determines Tissue Organization
Demeter, Derényi, Szöllősi Nature Communications 2022 https://www.nature.com/articles/s41467-022-29004-1
In this paper, we demonstrate that tissue structures are shaped by a balance between minimizing mutation accumulation and enhancing differentiation commitment, providing insights into how tissues might have evolved to mitigate cancer risks.
We root the tree of bacterial without using an out-group using genome-scale phylogenetic reconciliationand show that the last bacterial common ancestor was a free-living, flagellated, rod-shaped organism with a double membrane.
