Personalized logical models to investigate cancer response to BRAF treatments in melanomas and colorectal cancers¶

Installation¶

In order run the following notebook you will need this libraries: * pip install maboss * pip install maboss_test

[1]:

import maboss
import maboss_test

GINsim Model:¶

[2]:

from IPython.display import Image
Image("Model.png")

[2]:

Simulation of the wild tipe model¶

import the model the MaBoSS version of the model:

[3]:

BRAF_model = maboss.load("BRAF_Model.bnd", "BRAF_Model.cfg")

We simulate the model without any mutation and plot the probability to have the level of proliferation active:¶

[4]:

WT_model = BRAF_model.copy()
WT_model.network.set_output(('Proliferation_b1','Proliferation_b2'))

results_WT_model = WT_model.run()
# Plot pie chart
results_WT_model.plot_piechart()

Now we show again the same simulation but with BRAF mutated¶

[5]:

BRAFV600E_model = BRAF_model.copy()
BRAFV600E_model.mutate('BRAF', 'ON')
BRAFV600E_model.network.set_output(('Proliferation_b1','Proliferation_b2' ))

results_BRAFV600E_model = BRAFV600E_model.run()
# Plot pie chart
results_BRAFV600E_model.plot_piechart()

Automatized model-checking within unit-testing framework¶

In this notebook we test the consistence of our model through the unit-testing framework we developed.

title

Prepare the tool:¶

(for a brief tutorial of the tool go at: https://maboss-test.readthedocs.io/en/latest/?badge=latest)

Create the objects to perform the tests¶

We build two different objects, one corresponding to the network which presents the mutated form of BRAF (BRAF_ON) and the other for the treated network (BRAF_OFF).

[6]:

BRAFV600E = maboss.load("BRAF_Model.bnd", "BRAF_Model.cfg")

BRAFV600E.mutate('BRAF', 'ON')
test_BRAF_ON = maboss_test.MaBoSSTestCase(BRAFV600E)   #tester for the BRAF mutated network

BRAFV600E.mutate('BRAF', 'OFF')
test_BRAF_OFF = maboss_test.MaBoSSTestCase(BRAFV600E)  #tester for the treated network

Definition of some useful variables:¶

As we mentioned in the article, CRC and melanoma present different level of expression of SOX10. We define the intial conditions:

[7]:

IC_CRC = {'SOX10':[1,0]}    #initial conditions for CRC: SOX10 not expressed
IC_Mela = {'SOX10':[0,1]}   #initial conditions for melanoma: SOX10 expressed

Variables to simulate different kind of treatment:

[8]:

BRAF_inhibition = {'BRAF':'OFF'}
MEK_inhibition =  {'MEK_b1':'OFF', 'MEK_b2':'OFF'}
BRAF_EGFR_inhibition = {'BRAF' :'OFF', 'EGFR':'OFF'}
BRAF_MEK_PI3K_inhibition = {'BRAF':'OFF', 'MEK_b1':'OFF', 'MEK_b2':'OFF', 'PI3K':'OFF'}

Assertions list:¶

Notice: in the following tests the initial conditions are random for every node (each node starts with a 50% probability to be active), except when they are specifically changed (second argument of the function).

Assertion 1¶

source: doi:10.1038/nature10868
assertion: BRAF inhibition causes a feedback activation of EGFR in colorectal cancer andnot in melanoma.

[9]:

test_BRAF_ON.assertStateProbabilityEvolution(BRAF_inhibition, IC_CRC, {'EGFR':1}, 'increase', digits=1)

True!
The new probability of reaching the state is: 0.278416
The old one is: 0

[10]:

test_BRAF_ON.assertStateProbabilityEvolution(BRAF_inhibition, IC_Mela, {'EGFR':1}, 'stable', digits=1)

True!
The new probability of reaching the state is: 0.035271
The old one is: 0

The probability to have the riactivation of EGFR is higher in CRC than in melanoma.

Assertion 2¶

source: doi: 10.1158/0008-5472.CAN-10-0902, doi 10.1007/s00018-016-2297-8
assertion: MEK inhibition stops ERK signal but activates the PI3K/Akt pathway and increases the activity of ERBB3.

MEK inhibition stops ERK signal

[11]:

test_BRAF_ON.assertStateProbabilityEvolution(MEK_inhibition, {}, {'ERK_b1':1}, 'decrease')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0
The old one is: 1.0

[12]:

test_BRAF_ON.assertStateProbabilityEvolution(MEK_inhibition, {}, {'ERK_b2':1}, 'decrease')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0
The old one is: 1.0

but activates the PI3K/Akt pathway

[13]:

test_BRAF_ON.assertStateProbabilityEvolution(MEK_inhibition, {}, {'PI3K':1,'AKT':1}, 'increase')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0.462
The old one is: 0.2456

and increases the activity of both EGFR and ERBB3.

[14]:

test_BRAF_ON.assertStateProbabilityEvolution(MEK_inhibition, {}, {'EGFR':1}, 'increase')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0.259
The old one is: 0

Assertion 3¶

source: doi:10.1038/nature11183.
assertion: HGF signal leads to the reactivation of the MAPK and PI3K/AKT pathways, and resistance to BRAF inhibition.

[15]:

test_BRAF_OFF.assertStateProbabilityEvolution({'HGF':'ON'}, {}, {'PI3K':1,'AKT':1}, 'increase' )

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0.4946
The old one is: 0.337801

[16]:

test_BRAF_OFF.assertStateProbabilityEvolution({'HGF':'ON'}, {}, {'ERK_b1':1}, 'increase' )

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0.45903
The old one is: 0.431094

Assertion 4¶

source: DOI: 10.1038/s41467-017-02354-x
assertion: BRAF inhibition in melanoma activates the SOX10/FOXD3/ERBB3 axis, which mediates resistance through the activation of the PI3K/AKT pathway.

[17]:

axis = {'SOX10':1,'FOXD3': 1,'ERBB3':1}

[18]:

test_BRAF_ON.assertStateProbabilityEvolution(BRAF_inhibition, IC_Mela, axis, 'increase')

True!
The new probability of reaching the state is: 0.475832
The old one is: 0

[19]:

test_BRAF_ON.assertStateProbabilityEvolution(BRAF_inhibition, IC_Mela, {'ERBB3':1,'PI3K':1,'AKT':1}, 'increase')

True!
The new probability of reaching the state is: 0.241214
The old one is: 0

assertion: depletion of SOX10 would sensitize melanoma cells to the BRAF inhibitors.

[20]:

test_BRAF_OFF.assertStateProbabilityEvolution({'SOX10':'OFF'}, IC_Mela, {'Proliferation_b1':1}, 'decrease')

True!
The new probability of reaching the state is: 0.209514
The old one is: 0.244587

[21]:

test_BRAF_OFF.assertStateProbabilityEvolution({'SOX10':'OFF'}, IC_Mela, {'Proliferation_b2':1}, 'decrease')

True!
The new probability of reaching the state is: 0.118466
The old one is: 0.145325

Assertion 5¶

source: doi: 10.21037/atm.2016.06.07
source: Overexpression/mutation of CRAF results in constitutive activation of ERK and MEK also in the presence of a BRAF inhibitor.

In order to thest this assertion we build a simulation with both BRAF and CRAF constitutively active:

[22]:

BRAFV600E.mutate('BRAF', 'ON')
BRAFV600E.mutate('CRAF', 'ON')
test_BRAF_CRAF_ON = maboss_test.MaBoSSTestCase(BRAFV600E)

[23]:

test_BRAF_CRAF_ON.assertStateProbabilityEvolution(BRAF_inhibition, {}, {'MEK_b1':1,'ERK_b1':1}, 'stable')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 1.0
The old one is: 1.0

Assertion 6¶

source: doi: 10.21037/atm.2016.06.07
assertion: Early resistance to BRAF inhibition may be observed in case of PTEN loss, or mutations in PI3K or AKT.

PTEN loss leads to an increase of proliferation in BRAF treated patients

[24]:

test_BRAF_OFF.assertStateProbabilityEvolution({'PTEN':'OFF'}, {}, {'Proliferation_b1':1}, 'increase')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0.425284
The old one is: 0.22255

[25]:

test_BRAF_OFF.assertStateProbabilityEvolution({'PTEN':'OFF'}, {}, {'Proliferation_b2':1}, 'increase')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0.263994
The old one is: 0.124011

PI3K or AKT activation lead to an increase of proliferation in BRAF treated patients

[26]:

test_BRAF_OFF.assertStateProbabilityEvolution({'PI3K':'ON'}, {}, {'Proliferation_b1':1}, 'increase')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0.499279
The old one is: 0.22436

[27]:

test_BRAF_OFF.assertStateProbabilityEvolution({'PI3K':'ON'}, {}, {'Proliferation_b2':1}, 'increase')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0.361824
The old one is: 0.129211

[28]:

test_BRAF_OFF.assertStateProbabilityEvolution({'AKT':'ON'}, {}, {'Proliferation_b1':1}, 'increase')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0.446788
The old one is: 0.228203

[29]:

test_BRAF_OFF.assertStateProbabilityEvolution({'AKT':'ON'}, {}, {'Proliferation_b2':1}, 'increase')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0.148955
The old one is: 0.135235

Assertion 7¶

source: doi: 10.21037/atm.2016.06.07
Experiments in melanoma cell lines support combined treatment with BRAF/MEK + PI3K/AKT inhibitors to overcome resistance.

[30]:

test_BRAF_ON.assertStateProbabilityEvolution(BRAF_MEK_PI3K_inhibition, {}, {'Proliferation_b1':1}, 'decrease')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0
The old one is: 0.5012

[31]:

test_BRAF_ON.assertStateProbabilityEvolution(BRAF_MEK_PI3K_inhibition, {}, {'Proliferation_b2':1}, 'decrease')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0
The old one is: 0.2508

Assertion 8¶

source: doi:10.1158/2159-8290.CD-11-0341.
BRAF inhibition (Vemurafenib) also leads to induction of PI3K/AKT pathway and inhibition of EGFR did not block this induction.

[32]:

test_BRAF_ON.assertStateProbabilityEvolution(BRAF_inhibition, {}, {'PI3K':1,'AKT':1}, 'increase')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0.355729
The old one is: 0.2486

[33]:

test_BRAF_ON.assertStateProbabilityEvolution(BRAF_EGFR_inhibition, {}, {'PI3K':1,'AKT':1}, 'increase')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0.337245
The old one is: 0.2546

‘’Despite the profound effect of this combination on cell viability’‘: the inhibition of both EGFR and BRAF when the EGF signal is strong is better than just BRAF inhibition

[34]:

test_BRAF_ON.assertStateProbabilityEvolution(BRAF_EGFR_inhibition, {'EGF':[0,1]}, {'Proliferation_b1':1}, 'decrease')

True!
The new probability of reaching the state is: 0.203593
The old one is: 0.5082

[35]:

test_BRAF_ON.assertStateProbabilityEvolution(BRAF_inhibition, {'EGF':[0,1]}, {'Proliferation_b1':1}, 'decrease')

True!
The new probability of reaching the state is: 0.228733
The old one is: 0.4988

Assertion 9:¶

source: doi:10.1158/2159-8290.CD-11-0341
assertion: Induction of PI3K-AKT pathway signaling has been associated with decreased sensitivity to MAPK inhibition.

[36]:

test_BRAF_OFF.assertStateProbabilityEvolution({'PI3K':'ON'}, {}, {'Proliferation_b1':1}, 'increase')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0.504934
The old one is: 0.223471

[37]:

test_BRAF_OFF.assertStateProbabilityEvolution({'PI3K':'ON'}, {}, {'Proliferation_b2':1}, 'increase')

Warning, you did not change any initial condition!
True!
The new probability of reaching the state is: 0.361837
The old one is: 0.143544

[ ]: