notebook.community

Edit and run



In [1]:

    
%pylab inline
matplotlib.rcParams['figure.figsize'] = (10,8)









    



Populating the interactive namespace from numpy and matplotlib



In [2]:

    
from cno import cnodata, CNORbool



In [3]:

    
c = CNORbool(cnodata("PKN-ToyMMB.sif"), cnodata("MD-ToyMMB.csv"))
#c = CNORbool(cnodata("PKN-ExtLiverPCB.sif"), 
#             cnodata("MD-ExtLiverPCB.csv"))
#c = CNORbool(cnodata("PKN-LiverDREAM.sif"), 
#             cnodata("MD-LiverDREAM.csv"))



In [4]:

    
c.cnograph.png









    Out[4]:

Optimisation

Done with a Genetic Algorithm



In [5]:

    
# the default exammple takes about 5-10 seconds
c.optimise()









    



{'reltol': 0.1, 'pmutation': 0.5, 'maxstallgens': 100, 'time_index_1': 1, 'popsize': 50, 'elitism': 5, 'selpress': 1.2, 'sizefactor': 0.0001, 'maxgens': 500, 'maxtime': 60, 'NAFac': 1, 'verbose': True}

Warning in Models. found a + sign... in EGF+TNFa=PI3K. Interepreted as ^
Warning in Models. found a + sign... in Raf+!Akt=Mek. Interepreted as ^
Warning in Models. found a + sign... in Erk+TNFa=Hsp27. Interepreted as ^



In [6]:

    
c.results.results.keys()









    Out[6]:





['reactions',
 'models',
 'all_scores',
 'stimuli',
 'all_bitstrings',
 'results',
 'best_score',
 'midas',
 'species',
 'sim_results',
 'best_bitstring',
 'pkn',
 'inhibitors']

The optimal model and score

In the Genetic Algorithm, we have actually a set of models stored as well as the best model



In [7]:

    
c.results.results['best_bitstring'],c.results.results.reactions









    Out[7]:





(array([1, 1, 0, 1, 1, 0, 1, 1, 0, 0, 1, 1, 1, 0, 0, 0]),
 ['EGF=PI3K',
  'TNFa=PI3K',
  'Jnk=cJun',
  'PI3K=Akt',
  'Raf=Mek',
  '!Akt=Mek',
  'Mek=p90RSK',
  'Mek=Erk',
  'Erk=Hsp27',
  'TNFa=Jnk',
  'TNFa=NFkB',
  'TNFa=Hsp27',
  'EGF=Raf',
  'EGF^TNFa=PI3K',
  '!Akt^Raf=Mek',
  'Erk^TNFa=Hsp27'])



In [8]:

    
c.models.scores.min()









    Out[8]:





0.0296426065162907



In [9]:

    
c.models.df.ix[0:2]









    Out[9]:






  
    
      
      EGF=PI3K
      TNFa=PI3K
      Jnk=cJun
      PI3K=Akt
      Raf=Mek
      !Akt=Mek
      Mek=p90RSK
      Mek=Erk
      Erk=Hsp27
      TNFa=Jnk
      TNFa=NFkB
      TNFa=Hsp27
      EGF=Raf
      EGF^TNFa=PI3K
      Raf^!Akt=Mek
      Erk^TNFa=Hsp27
    
  
  
    
      0
       1
       1
       0
       1
       1
       0
       1
       1
       0
       0
       1
       1
       1
       1
       0
       0
    
    
      1
       1
       1
       0
       1
       1
       0
       1
       1
       0
       0
       1
       1
       1
       1
       0
       1
    
    
      2
       1
       1
       1
       1
       1
       0
       1
       1
       0
       0
       1
       1
       1
       0
       1
       0



In [10]:

    
c.models.cnograph.midas = c.midas  # could be done automatically in the code
c.models.cnograph.png









    Out[10]:



In [11]:

    
c.models.errorbar()



In [12]:

    
c.models.heatmap()









    Out[12]:





<biokit.viz.heatmap.Heatmap at 0x4729f90>

Report



In [13]:

    
# a new page should be openned automatically in a few seconds
c.onweb()









    



openning report_bool/index.html

simulate



In [14]:

    
c.simulate(c.results.results.best_bitstring)









    Out[14]:





0.0296426065162907

This function, which is actually a method could be used to solve the problem with any optimisation package.



In [15]:

    
def objfunc(parameters):
    return c.simulate(parameters)

objfunc([0,0,0,0, 0,0,0,0, 0,1,1,1, 1,1,1,1])









    Out[15]:





0.132822472848789



In [ ]: