Abstract
From microbial communities to cancer cells, many such complex collectives embody emergent and self-organising behaviour. Such behaviour drives cells to develop composite features such as formation of aggregates or expression of specific genes as a result of cell-cell interactions within a cell population. Currently, we lack universal mathematical tools for analysing the collective behaviour of biological swarms. To address this, we propose a multifractal inspired framework to measure the degree of emergence and self-organisation from scarce spatial (geometric) data and apply it to investigate the evolution of the spatial arrangement of Enterobacter cloacae aggregates. In a plate of semi-solid media, Enterobacter cloacae form a spatially extended pattern of high cell density aggregates. These aggregates nucleate from the site of inoculation and radiate outward to fill the entire plate. Multifractal analysis was used to characterise these patterns and calculate dynamics changes in emergence and self-organisation within the bacterial population. In particular, experimental results suggest that the new aggregates align their location with respect to the old ones leading to a decrease in emergence and increase in self-organisation.
The Challenge of Quantifying Collective Behavior
Biological systems, from bacterial swarms to metastatic cancer cells, exhibit complex collective behaviors that are greater than the sum of their individual parts. While we can observe these "emergent" properties—like the formation of specific spatial patterns—we lack universal mathematical tools to quantify them, especially when tracking every individual cell is impossible.
Most existing models require high-resolution microscopic data, which is often difficult to obtain in real-world experiments. This research introduces a "macro-perspective" method to measure how a system organizes itself over time using only the geometric distribution of cell aggregates captured in simple time-lapse photography.
Methodology: A Multifractal Framework
The researchers utilized Multifractal Analysis, a technique typically used to describe complex, self-similar objects (like coastlines or clouds) that look similar at different scales. Unlike a simple monofractal, a multifractal system possesses a spectrum of scaling exponents, reflecting varying densities across different scales.
To account for the circular Petri dish, the team used Radial Box-Counting to calculate the probability of finding an aggregate at various scales. From this, they derived two novel metrics:
- Emergence (E): The rate of change of the generalized dimensions over time.
- Self-Organisation (S): Based on the width and shape of the multifractal spectrum, representing the system's move toward an ordered state.
Experimental Observations in Enterobacter cloacae
The framework was applied to Enterobacter cloacae grown in semi-solid agar. The bacteria initially expand in a "swarm band" and then settle into dense, millimeter-scale aggregates. These aggregates nucleate from the site of inoculation and radiate outward to fill the plate.
Key Insights and Results
- Emergence is Finite: The emergence metric ($E$) peaks early and then drops to zero well before the Petri dish is fully populated. This suggests that once the "logic" of the pattern is established, adding more aggregates doesn't add new complexity to the system's organization.
- The Monofractal "Goal": As the bacteria self-organize, the multifractal spectrum shifts and narrows. The authors suggest the community is trying to reach a "monofractal" state—a state of perfect repetition and symmetry—to potentially balance cooperative and competitive interactions for limited nutrients.
- Spatial Memory: New aggregates don't appear randomly. Their positions are regulated by existing aggregates (likely via chemoattractant gradients), ensuring the system maintains a constant entropy.
- Independence from Density: Interestingly, the degree of self-organization did not correlate directly with the number of aggregates or the growth rate. This proves the metric captures the logic of the arrangement, not just the quantity of cells.
Conclusion and Future Directions
This study demonstrates that bacterial aggregates transition from high-emergence states to highly self-organised, stable patterns that mimic monofractal distributions. While currently limited to 2D patterns, this framework provides a powerful tool for quantifying collective behavior without needing to track individual cells. Future applications could include studying cancer metastasis or programming synthetic microbial communities to form specific functional structures.
Summary updated