Skip to main content

Information Processing in Biological Systems

Physics

Image of Gasper Tkacik

Gasper Tkacik

Professor


1 Career

  • Since 2022 Deputy Dean of the Graduate School, Institute of Science and Technology Austria (ISTA)
  • Since 2017 Professor, Institute of Science and Technology Austria (ISTA)
  • 2015 – 2022 Program Chair of the Graduate School, Institute of Science and Technology Austria (ISTA)
  • 2011 – 2017 Assistant Professor, Institute of Science and Technology Austria (ISTA) 
  • 2008 – 2010 Postdoc, University of Pennsylvania, Philadelphia, USA
  • 2007 PhD, Physics Department, Princeton University, Princeton, USA

1.1 Selected Distinctions

  • 2024 Physics of Cell Fate Decision, conference organizer (ISTA)
  • 2019 Lieben Prize of the Austrian Academy of Sciences

1.2 Selected Grants

  • 2026 – 2030 FWF-ANR (PIN4502925), Deciphering the cellular code for developmental time. Co-PI: Wolfgang Keil (Curie Institute). Awarded to RG: € 386,012.
  • 2024 – 2030 ERC Synergy (101118866), DynaTrans. Co-PIs: Thomas Gregor (Pasteur Institute), Denis Duboule (College de France). Awarded to RG: € 2,123,363.
  • 2024 – 2028 WWTF (LS23-026), Understanding pancreas biology with AI/ML. Co-PI: Marjan Slak-Rupnik (MedUniWien). Awarded to RG: € 227,697.

1.3 Tech Transfer (Patents, Spin-Offs)

N/A


1.4 Selected Publications

  1. Sokolowski TR et al., “Deriving a genetic regulatory network from an optimization principle,” Proc Natl Acad Sci USA 122: e240295121 (2025). https://doi.org/10.1073/pnas.2402925121.
  2. Ruzickova N et al., “Quantitative omnigenic model discovers interpretable genome-wide associations,” Proc Natl Acad Sci USA 121: e2402340121 (2024). https://doi.org/10.1073/pnas.2402340121.
  3. Reinhardt M et al., “Path Weight Sampling: Exact Monte Carlo computation of the mutual information between stochastic trajectories,” Phys Rev X 13: 041017 (2023). https://doi.org/10.1103/PhysRevX.13.041017.
  4. Hledik M et al., “Accumulation and maintenance of information in evolution,” Proc Natl Acad Sci USA 119: e2123152119 (2022). https://doi.org/10.1073/pnas.2123152119.
  5. Mlynarski WF et al., “Statistical analysis and optimality of neural systems,” Neuron 109: 1 (2021). https://doi.org/10.1016/j.neuron.2021.01.020.

2 Group Members

  • Reka Borbely, PhD Student since 2019
  • Camila Bräutigam, PhD Student since 2021 (co-advised with Nicholas Barton)
  • Athina Diakogianni, PhD Student since 2021 (co-advised with Calin Guet)
  • Ekaterina Maksimova, PhD Student since 2020 (co-advised with Matthew Robinson)
  • Pablo Mateu Hoyos, PhD Student since 2024 (co-advised with Edouard Hannezo)
  • Anuj Patel, PhD Student since 2022 (co-advised with Maximilian Jösch)
  • Tetiana Rabiichuk, PhD Student since 2022
  • Julian Renaud, PhD Student since 2022
  • Puzhen Xia, PhD Student since 2022 (co-advised with Calin Guet)
  • Bahti Zakirov, PhD Student since 2021
  • Tereza Zuskinova, PhD Student since 2024
  • Alexis Benichou, Postdoc since 2025
  • Zhang Alex Chen Yi, Postdoc since 2025
  • Iyer Krishnan, Postdoc since 2024
  • Gabor Lengyel, Postdoc since 2025
  • Elia Mascolo, Postdoc since 2026
  • Natalia Ruzickova, Postdoc since 2025

3 Research

We study how biological systems process information across scales, from molecular networks to neural populations and collective behaviors. By combining tools from statistical physics, information theory, and biophysics, we seek quantitative principles that explain how living systems reliably compute, sense, and make decisions despite noise, limited resources, and physical constraints.

Our research focuses on how network architecture, stochasticity, and evolutionary pressures shape information transmission and function. Key domains include gene regulation during development, neural coding in sensory systems, and the collective dynamics of interacting agents. A central goal is to link microscopic mechanisms—such as molecular interactions—to emergent system-level behaviors and their evolutionary consequences.

We work at the interface of theory and experiment, pairing mathematical modeling with high-precision data in close collaboration with experimental partners. By developing predictive, data-driven frameworks, we aim to uncover universal principles governing biological computation and explain how complex regulatory and neural networks achieve robust, efficient information processing.


3.1 Current Scientific Projects/Directions

  • Information theory for early developmental processes
  • Mechanisms of eukaryotic and prokaryotic gene regulation
  • Evolution of gene regulatory sequences and architecture
  • Adaptive coding and collective representations of stimulus in the retina, cortex and beyond
  • Critical and near-critical states of neural dynamics
  • Bayesian computations in the brain
  • Collective function in pancreatic islets of Langerhans for efficient blood glucose level control

3.2 Research in Focus

The paper “Deriving a genetic regulatory network from an optimization principle” studies how a complex molecular network in the fruit fly embryo can be explained from first principles. Instead of fitting a model to data, as in traditional approaches, we assume that evolution has optimized the network for a specific task: enabling each cell to infer its position from noisy molecular signals. This can be formalized by maximizing “positional information,” a long-standing biological concept that we previously cast in information-theoretic terms, allowing it to be computed or estimated from data. We then construct a detailed model of the “gap gene” system (with 50+ parameters) and optimize it under realistic physical constraints, such as limited molecule numbers. The resulting networks closely match the real system in both structure and gene expression patterns.

The key novelty is deriving a complex, mechanistic gene regulatory network from an explicit optimization principle rather than fitting it to data. In this sense, the work parallels theoretical physics approaches, where fundamental principles yield quantitative predictions about complex systems without relying primarily on statistical inference. Extending this reasoning to a realistic, high-dimensional biological network suggests that biological organization may reflect underlying physical or informational constraints. Crucially, it also provides a new lens on evolutionary questions: while earlier evo-devo studies often relied on toy models to explore concepts like necessity versus contingency, here these ideas can be investigated within a detailed, biologically grounded model that can either be optimized (mimicking evolution) or fit directly to data.

Reference:  Sokolowski TR et al., “Deriving a genetic regulatory network from an optimization principle,” Proc Natl Acad Sci USA 122: e240295121 (2025). https://doi.org/10.1073/pnas.2402925121.



theme sidebar-arrow-up
Back to Top