Hausdorff School: “Inverse problems for multi-scale models”

Dates: August 22 - 26, 2022

Venue: HIM lecture hall (Poppelsdorfer Allee 45, Bonn)

Organizers: Lorenzo Contento, Jan Hasenauer, Yannik Schälte

 

You can find the schedule here.

 

The key aim of mathematical, computational and systems biology is to achieve a holistic understanding of biological systems. Over decades, this aim was mainly approached by moving from the study of single molecules to the analysis of biological networks. However, it has been recognised that the study of cellular networks is only the first step, as multicellular organisms are more than just a collection of pathways. A mechanistic understanding of complex biological functions requires the consideration of multiple spatial and temporal scales. Multi-scale models for a variety of processes have been developed, such as the beating of the heart, the development of cancer, and drug delivery and action. A single multi-scale model usually requires the application of multiple mathematical techniques. For example, stochastic descriptions of individual cells can be coupled with continuous fields for extracellular substances, thereby connecting branching processes with partial differential equations.

In this Hausdorff School we will focus on the estimation of the parameters of multi-scale models from experimental data. The inverse problem of determining parameter values from experimental data is mathematically and computationally challenging. We will introduce the participants to advanced likelihood-based and -free statistical inference methods, including state-of-the-art Approximate Bayesian Computing (ABC) methods. Furthermore, novel strategies for the reduction of the computational complexity will be discussed, including multi-resolution and surrogate modelling.

This School is intended for motivated graduate students, PhD candidates or postdoctoral researchers in mathematical modelling. The mini-courses given by the key speakers will be complemented by individual talks by selected participants, tutorial sessions on relevant software tools (e.g., the modelling framework Morpheus and the parameter estimation toolbox pyABC) and plenty of opportunities for scientific exchange and networking.


Key Speakers: The following speakers will give a lecture series

  •  Brian Drawert (Asheville)
  •  Stefam Engblom (Uppsala)
  •  Christiane Fuchs (Bielefeld)
  •  Andreas Hellender (Uppsala)
  •  Linda Petzold (Santa Barbara)
  •  Dennis Prangle (Bristol)

 

The application to participate in this Hausdorff School has been closed.