Research Vision

Predictive Evolutionary Biology: Unveil the rules of biological system evolution through predictive modeling

My long-term research interest lies in uncovering the rules that govern the evolution of complex biological systems. To this end, I have been exploring the predictability of evolution—a long-standing question in biology with broad implications not only for evolutionary biology but also for medicine and strategic bioengineering. 

I have contributed to understanding the predictability of complex biological system evolution by developing novel computational frameworks and uncovering recurrent evolutionary patterns. Currently, there are three remarkable challenges for predicting evolution: (1) scalability of past evolution inference, (2) lack of computational framework to predict evolution, and (3) lack of model studies unveiling novel evolutionary patterns. 

(1) Scalable evolutionary inference. I developed FRACTAL, a parallelized phylogenetic inference method enabling high-resolution tree reconstruction from large-scale genomic data (Nat. Biotechnol., 2022), which could also contribute to cell lineage tracing in developmental biology (Science, 2022). 

(2) Forecasting framework for evolution. I established Evodictor, a predictive model for evolution by gene gain and loss in prokaryotic genomes, revealing the predictability of metabolic evolution and the underlying rules (Sci. Adv., 2023).  I'm currently extending the Evodictor framework for application to diverse study targets.

(3) Previously unrecognized evolutionary patterns. I studied recurring evolution in the past, such as convergent gene loss order in lactic acid bacteria (Commun. Biol., 2024) and protein structural convergence resulting from independent gene fusions (bioRxiv, 2025). To consider ecological aspects, my work has further explored causal relationships between habitat breadth and functional evolution (bioRxiv, 2024) and patterns behind horizontal gene transfers across plasmids with diverse host ranges (bioRxiv, 2025).

Predictive evolutionary biology seeks more than just forecasting evolution. By identifying patterns in past evolutionary events, we can infer functional relationships between genes, thereby enabling the prediction of functions for uncharacterized genes. For example, I predicted and experimentally validated the function of an uncharacterized gene involved in alcohol metabolism (bioRxiv, 2025), and investigated genes critical for microbial adaptation to diverse light environments. Through these efforts and international collaborations, I continue to broaden the scope of my research.

While I targeted microbial genomic and molecular evolution in my Ph.D., I'm now expanding my research to multicellular animals. At Stanford University, I start studying the evolutionary constraints and predictability of cell differentiation processes during development.