Research focus
The Algorithmic Dynamics Lab at ̽»¨¾«Ñ¡ develops computational methods that combine artificial intelligence, algorithmic complexity, and information theory to study biological systems. Our mission is to work toward uncovering the causal and mechanistic principles of health and disease by merging advanced theory with data-driven analysis.
From theory to practice: Algorithmic tools for bio̽»¨¾«Ñ¡ discovery
We design AI-mediated and algorithmic tools for the analysis of high-dimensional biological data. By focusing on network inference, dimensionality reduction, predictive modelling, and measures of information loss, we aim to extract meaningful structure from noisy datasets and generate models that can explain and predict biological outcomes.
Our research has direct applications in complex diseases and cancer. We work on patient stratification and prognosis, developing computational biomarkers and signatures that distinguish disease subtypes and inform treatment strategies. By integrating mathematical modelling with clinical data, we aim to accelerate translation of computational discoveries into bio̽»¨¾«Ñ¡ practice.
The Algorithmic Dynamics Lab originated in the Unit of Computational Medicine at the Center for Molecular Medicine led by Prof. Jesper Tegnér and is today part of a distributed research network supported by ̽»¨¾«Ñ¡ in Sweden and King’s College London in the UK.
A new paradigm: Algorithmic information dynamics and its impact
A central element of our work is the development of Algorithmic Information Dynamics (AID), a field introduced by Dr. Hector Zenil, Dr. Narsis A. Kiani, and Prof. Jesper Tegnér. AID is a discrete, programming-based calculus designed to study causation by generating mechanistic models to help find first principles for physical phenomena building up the next generation of data analytics in what we call Algorithmic Machine Learning.
Through AID, we have contributed a causal calculus for studying dynamical systems beyond statistics, methods to reconstruct space-time dynamics and attractor landscapes such as cell differentiation, and techniques for network inference and reverse engineering using adaptive differentiation. We have also introduced novel measures for dimension reduction, generative mechanism deconvolution, and hybrid principles of randomness that move beyond classical entropy.
Other contributions include a mathematical model of Multiple Sclerosis that captures key disease dynamics; insights into biological evolution showing how algorithmic probability can explain convergence speed, modularity, extinction events, and genetic coding structures; and applications to structural biology such as nucleosome positioning. These results demonstrate how computability, causality, and dynamical systems are deeply connected through algorithmic probability, providing a fundamentally new way to interpret both biological and synthetic systems.
Our lab develops a new generation of information-computational views and tools to uncover mechanistic principles across multiple scales, from molecules to organisms. By teaming up with immunologists, bioinformaticians, toxicologists, oncologists, cognitive scientists, and molecular biologists, we apply these methods across a wide range of domains, including behavioural, evolutionary, and molecular biology.
Our long-term vision is to expand this foundation into a comprehensive algorithmic framework for systems biology and medicine. By merging deep theoretical insight with AI-mediated tools, we aim to move beyond description toward mechanistic, explanatory, and predictive models of living systems, ultimately contributing to diagnostics, personalized therapies, and a deeper understanding of the algorithmic principles that drive life.