This wiki service has now been shut down and archived
Stochastic effects in microbial infection
The workshop on stochastic effects in microbial infection was held at eSI September 28th-29th 2010, with a smaller team meeting on 30th September to summarise the key ideas. The lead organiser was Rosalind Allen. The workshop involved 21 participants, including microbiologists, evolutionary biologists, engineers, mathematicians, physicists and computer scientists and consisted of 19 talks followed by a roundtable discussion. 7 of the speakers were international.
Key ideas emerging from this workshop were:
The need for clear definitions of important concepts, to allow modellers and microbiologists to communicate better. Concepts with the potential to cause confusion include the definition of infection, the root causes of stochastic variation in populations and the difference between noise, bistability, bimodality and phase variation.
The need to determine the biological role of population heterogeneity. Microbiologists have long focussed on identifying the molecular mechanisms underlying the observed heterogeneity in microbial populations. Mathematical and computational scientists on the other hand have constructed very simplistic models which test the hypothesis that random switching can be beneficial in different scenarios. However the “middle ground” is lacking: experimental investigations of the role of switching, or conversely biologically realistic models for population dynamics of infections.
The need to disentangle the effects of growth, switching and environmental changes. Experimental observations usually consist of a “snapshot” of a microbial population: one must then work out what components of the observed population distribution are due to growth, due to random switching mechanisms or due to environmental changes undergone by the population during the experiment. This is a difficult task which may be overcome by sophisticated experiments with labelled microbes, in combination with mathematical / computational models. How random is “random switching”? In computational models, it is often assumed that microbial cells switch randomly between two states, while the environment also switches randomly between states. However, in many cases, cells actually modify their rate of switching in response to the environment. There is thus a continuum of possibilities ranging from completely random switching to completely “responsive” switching; this should be taken into account in models.
The need for large-scale computations connecting intracellular genetic network dynamics (simulated with stochastic algorithms such as the Gillespie Algorithm) with population dynamics. This would allow one to determine the effects of molecular changes (eg mutations) on the progress of an infection, or the spatial structure of a microbial community. This is a task which poses new computational challenges.
These ideas are explored in more detail in the roadmap document http://www2.ph.ed.ac.uk/~rallen2/infection_report.pdf.