Lunch Pitch with Lucas Hedström and Alexis Sullivan

Lunch Pitches with Lucas Hedström and Alexis Sullivan

Pitch 1: Lucas Hedström: What is the role of mulifractals in living systems?

Doctoral student at Department of Physics

 

Abstract:

Fractals are mathematical patterns that self-repeat at different scales. We find them in many natural systems, including the pattern of snowflakes, the structure of tree branches, and the shape of leaves. The repeating patterns of fractals are not just nice to look at; they also serve important functions in biology. For example, blood vessel branching ensures that each cell in the body receives the nutrients it needs to survive. On small scales, we can see fractal structures in how our genetic code, our DNA, is folded in our cells. DNA folding plays a critical role in processes like cell division and gene regulation. Failure in these processes can result in harmful conditions such as cancer.

To better understand how fractal structures influence biological processes, we want to discuss with people who know more about fractal structures and biological systems where fractals occur.

 

Imagine your favorite mushroom hunting spot. Is porcini and chanterelle risotto in your future, or will you only find enough for a slice of toast? Predicting the coming autumn’s bounty, surprisingly, is analogous to sending astronauts to the moon. In both cases, the value of interest, be it abundance or trajectory, are inferred from imprecise measurements. Independently, we reasonably expect their current, unknown values to be linked to their previous states: in engineering parlance, they are dynamical systems, or in other words, are autocorrelated. With repeated observations, predictions based on previous observations and an autocorrelation function can be compared to incoming data and any discrepancies used to refine future predictions. These elements – a recursive algorithm, set of observations, and a dynamic process – comprise a state space model, which made their grand debut with the Apollo 11 lunar landing. Some decades later, these models are increasingly applied to the messy data often collected by ecologists.

In this pitch, I will share insights gained from studying the dynamics of ca. 2,000 organisms in a 34-year environmental DNA archive. While an improvement over previous methods, I believe we have yet to realize the full potential state space models in ecology. This is due, in part, because the decades of progress made by engineers and physicists are effectively inaccessible to outsiders. Whether you’re fluent in measure theory or have ever cursed your noisy, autocorrelated data, I’m looking for your help to design, derive, and develop the tools to launch ecology’s moonshot moment.