|Current Workshop | Overview||Back | Home | Search ||
Multiscale Systems Biology of Cancer
Aim of the workshop
Although cancer is typically seen as a disease of the genes, in fact many phenomena, including tumor plasticity, metastasis, and relapses after therapies, can only be understood if we look beyond the molecules, e.g. at individual cell behavior, cell-cell competition, cell-stroma interactions, and metabolism. Collaborations between experimental and computational biologists are key to unraveling these multilevel interplays. To this end, experimental and computational researchers working at different ends of the multiscale spectrum discussed with one another at this Lorentz workshop. Scales range from the molecular level, the cellular and tissue level, to physiology and public health genomics. The workshop centered around plenary discussions on how to couple different spatial and time scales from experimental, computational, and philosophical viewpoints. The workshop encouraged participants to think “out of their scale” such that they will apply this in their research.
Throughout the meeting, extensive notes were taken of the discussions, as input for a white paper that will become a tangible outcome of the workshop. Also, new collaborations on multiscale modeling were set up; a particularly interesting new direction here are projects to model the feedback between tissue structure and efficiency of metabolism in tumors and in the liver.
Scientific breakthroughs and ‘Aha’ moments
A particularly interesting discussion emerged between two apparent “philosophical approaches” to multiscale modeling in biology. Roughly, one approach views the scales in a multicellular organism (genes, molecules, cells, tissues) as an inseparable continuum, that are all equally important for fully explaining any biological phenomenon. In this view, the ideal model would describe the whole organisms in terms of all its molecules, but the model would become too large to compute in reasonable time. Thus in this view, multiscale modeling is primarily a technological challenge to make whole-organism simulations feasible and tractable. An alternative vision is also known as the “middle-out” approach (Noble, The Music of Life; attributed to Brenner). Here a “central scale” is chosen depending on the type of question asked: e.g., the molecular network to ask questions about genetic regulation, the tissue scale to ask questions about biomechanics or physiology. Then, depending on what is needed to explain the phenomenon, relevant detail from the smaller and larger scales will be taken into account in the description, e.g., cellular structure or ion channels in a heart model, or feedback from adjacent cells in a model of a genetic regulatory network. Here, the scales are often chosen based on biological structure, and make use of the inherent, modular, nested structure of biological organisms. Thus, here a multiscale model is seen as a natural representation of an inherently multiscale phenomenon. It was rewarding to see how during the workshop representatives of the two views started to understand and appreciate each other’s stances more, and how nevertheless the different views lead to the same modeling choices. Another interesting development in the workshop was to see how experimental biologists use new computational modeling insights to design new experiments and to develop new experimental concepts.
Organization/format of the workshop
In this workshop there were very few formal talks talks planned. Instead, most of the time was devoted to plenary discussion, and opportunities for last-minute talks. This format worked very well for us and give a very dynamic workshop atmosphere. We were with a relatively small group, and after the first day we moved to the Gratama room. The less formal set-up in that room encouraged all participants, including many students, to participate actively in the discussions.
Roeland Merks (Amsterdam, Netherlands)
Matthias Reuss (Stuttgart, Germany)
Hans Westerhoff (Manchester, United Kingdom)