A Mechanical Model of Rod-Shaped Bacteria

## A Mechanical Model of
## Rod-Shaped Bacteria
#### Jonas Pleyer
##### 30.04.2025

---

## Goal: Model this System

---

### Rod-like bacterial shape is maintained by feedback
### between cell curvature and cytoskeletal localization
#### (Ursell et al. 2014)

---

### Bending forces plastically
### deform growing bacterial cell walls
#### (Amir et al. 2014)

---

### Controlling the Shape of Filamentous
### Cells ofEscherichia coli.
#### (Takeuchi et al. 2005) *E.Coli*

---

### From Cell Differentiation to Cell Collectives:
### Bacillus subtilis Uses Division of Labor to Migrate
#### (Gestel et al. 2015)

---

### (C) Cellular Aspects

| Aspect | Description |
| --- | --- |
| (1) Rod-Shaped Mechanics | Rod-shaped bacteria are (flexible)  (Takeuchi et al. 2005, Ursell et al. 2014, Amir et al. 2014) |
| (2) Growth | Cells grow exponentially (Takeuchi et al. 2005, Robert et al. 2014). |
| (3) Differentiation | B.subtilis differentiates into matrix-producing and surfactin- producing cells (López and Kolter 2010, Gestel et al. 2015). |
| (4) Division | Formation of van Gogh bundles is driven by cell division (Gestel et al. 2015). Bacteria can form multilayers during their growth phase (Duvernoy et al. 2018). |
| (5) Variable Parameters | Parameters for individual cells are taken from a distribution (Koutsoumanis and Lianou 2013). |

---

### (CC) Cell-Cell Interactions

| Aspect | Description |
| --- | --- |
| (6) Adhesion (polar) | Bacteria adhere to each other and attach when in close contact (Verwey 1947, Trejo et al. 2013). The interaction can be polarized (Duvernoy et al. 2018). |
| (7) Friction | Friction between cells (Grant et al. 2014) can be asymmetrical (Doumic et al. 2020). |

<br>

### (DC) Domain-Cell Interactions

| Aspect | Description |
| --- | --- |
| (8) External Forces | Bacteria stick to surfaces (Loosdrecht et al. 1989) and interact with the extracellular (Grant et al. 2014). |
| (9) Extracellular Reactions | Bacteria can take up nutrients or possible secrete/take up signalling molecules (Li et al. 2025). |

---

### Our Current State

---

### Parameter Estimation
### With Microscopic Time Series

<div style="grid-template-columns: 1fr 1fr 1fr; display: grid; column-gap: 50px;">
<p>Initial</p>
<p>Final</p>
<p>Prediction</p>
<img width=600 src="figures/crm_fit/0012/snapshot-initial.png">
<img width=600 src="figures/crm_fit/0012/snapshot-final.png">
<img width=600 src="figures/crm_fit/0012/snapshot-final-predicted.png">
</div>

$\rightarrow$ Comparing predicted and actual positions allows us to calibrate the model parameters.

---

### Parameter Estimation
### With Microscopic Time Series

---

### Image Generation

<div style="grid-template-columns: 1fr 1fr 1fr; display: grid; column-gap: 0px;">
<p>Raw Format</p>
<p>Generate Microscopic Image</p>
<p>Generate Masks</p>
<img width=600 src="figures/generated/raw_pv/000000600.png">
<img width=600 src="figures/generated/images/000000600.png">
<img width=600 src="figures/generated/masks/000000600.png">
</div>

$\rightarrow$ Use dataset to train machine learning models

- cell-segmentation
- cell-tracking.

---

### Python Package

<pre class="stretch"><code data-trim data-noescape class="language-python">
import cr_mech_coli as crm

# Define Settings
config = crm.Configuration()
ag_set = crm.AgentSettings()

# Generate positions and agents
positions = crm.generate_positions_old(4, ag_set,config)
agents = [crm.RodAgent(
        pos=p,
        vel=p*0.0,
        **ag_set.to_rod_agent_dict()
    ) for p in positions
]

# Run simulation and plot individual results
cells = crm.run_simulation_with_agents(agents, config)
crm.store_all_images(cells, config.domain_size)
</code></pre>

</div>
<div>
<h3>Documentation</h3>
<iframe
    src="https://jonaspleyer.github.io/cr_mech_coli/"
    style="width: 1100px; height: 980px; max-width: 100%; max-height: 100%;"
></iframe></div>
</div>