[1]:

import maboss
import matplotlib.pyplot as plt

Analysis of the effect of scheduled treatments on a Cell fate Boolean model

In this Jupyter notebook, we’ll look at the properties of the model by Calzone et al. and study the effect of a scheduled TNF treatment on the possible cell fates.

Initializing the simulation

We start by initializing a simulation with the BND and CFG files of the model.

[12]:

bnd_file = "cellfate.bnd"
cfg_WT = "cellfate.cfg"

[13]:

param = {'max_time': 105, 'time_tick': 1, 'sample_count': 100000}
outputs = ['NonACD', 'Apoptosis', 'Survival']

[14]:

sim = maboss.load(bnd_file, cfg_WT)

Here, no TNF treatment

We start by inactivating the TNF input node, to simulate the effect of the absence of treatment

[15]:

sim.param.update(param)
sim.network.set_istate('TNF', [1, 0])
sim.network.set_output(outputs)

[16]:

res = sim.run()
fig = plt.figure(figsize=(3,3), dpi=500, tight_layout=True)
fig.subplots(1)
res.plot_piechart(axes=fig.axes[0])

../_images/notebooks_CellFateScheduling_8_0.png

We observe that in the lack of treatment, the cells will remain in the <nil> state, which corresponds to homeostasis.

Same one with TNF treatment

This time we activate the TNF node to simulate a treatment with TNF.

[17]:

sim_tnf = sim.copy()
sim_tnf.network.set_istate('TNF', [0, 1])
res_tnf = sim_tnf.run()
fig = plt.figure(figsize=(5,3), dpi=500)
fig.subplots(1)
res_tnf.plot_piechart(axes=fig.axes[0])

../_images/notebooks_CellFateScheduling_11_0.png

After the TNF treatment, we observe that the majority of the population undergo apoptosis (~83%) or necrosis (Non Apoptotic Cell Death, ~11%), while minority survives the treatment (~6%).

[18]:

fig = plt.figure(figsize=(5,3), dpi=500)
fig.subplots(1)
res_tnf.plot_trajectory(prob_cutoff=0, axes=fig.axes[0])

../_images/notebooks_CellFateScheduling_13_0.png

Looking at the trajectories, we see that the model reaches a steady state around t=30.

Simulating scheduled TNF treatments

Here we want to simulate the effect of successive TNF treatments, to see if those can improve the response.

To achieve this, we use the node scheduling functionality of MaBoSS.

[23]:

sim_tnf_schedule = sim_tnf.copy()
sim_tnf_schedule.network['TNF'].set_schedule({30: 0, 60: 1, 90: 0})

[24]:

res_tnf_schedule = sim_tnf_schedule.run()

[27]:

fig = plt.figure(figsize=(5,3), dpi=500)
fig.subplots(1)
res_tnf_schedule.plot_trajectory(prob_cutoff=0, axes=fig.axes[0])
fig.axes[0].axvspan(0., 30, facecolor='0.9')
fig.axes[0].axvspan(60, 90, facecolor='0.9')

[27]:

<matplotlib.patches.Rectangle at 0x7307fcc03890>

../_images/notebooks_CellFateScheduling_18_1.png

In this figure, we highlight in grey the TNF treatments.

We observe that after the first treatment, the cells which were surviving go back to homeostasis and become succet