Python Dynamic analysis of biochemical NetwOrks (PyDyNo) is an open source python library for the analysis of signal execution in network-driven biological processes. PyDyNo supports the analysis of PySB and SBML models.
PrePrint: Probability-based mechanisms in biological networks with parameter uncertainty
Oscar O. Ortega, Blake A. Wilson, James C. Pino, Michael W. Irvin, Geena V. Ildefonso, Shawn P. Garbett, Carlos F. Lopez
bioRxiv 2021.01.26.428266; doi: https://doi.org/10.1101/2021.01.26.428266
The preprint paper can be found here
Jupyter notebooks with the code to reproduce the paper figures can be found here
> pip install pydyno> pip install git+git:https://github.com/LoLab-VU/pydyno.git- Download and extract pydyno
- Navigate into the pydyno directory
- Install (Python is necessary for this step):
> python setup.py installimport pydyno
import numpy as np
from os.path import dirname, join
from IPython.display import Image
from pydyno.examples.double_enzymatic.mm_two_paths_model import model
from pydyno.visualize_simulations import VisualizeSimulations
from pydyno.discretization import PysbDomPath
from pydyno.visualize_discretization import visualization_path
from pysb.simulator import ScipyOdeSimulator# import calibrated parameters
module_path = dirname(pydyno.__file__)
pars_path = join(module_path, "examples", "double_enzymatic", "calibrated_pars.npy")
pars = np.load(pars_path)# define time for the simulation and simulate model
tspan = np.linspace(0, 100, 101)
sim = ScipyOdeSimulator(model, tspan=tspan).run(param_values=pars[:100])vt = VisualizeSimulations(model, sim, clusters=None)
vt.plot_cluster_dynamics(components=[5])
# This saves the figure in the local folder with the filename comp0_cluster0.png
dp = PysbDomPath(model, sim)
signatures, paths = dp.get_path_signatures('s5', 'production', depth=2, dom_om=1)
signatures.sequences.head()signatures.dissimilarity_matrix()
signatures.silhouette_score_agglomerative_range(4)# Select the number of cluster with highest silhouette score
signatures.agglomerative_clustering(2)# Plot signatures
signatures.plot_sequences()
# File is saved to the local directory with the filename modal.png
paths{2: [OrderedDict([('s5', [['s3'], ['s4']])]),
OrderedDict([('s3', [['s0', 's1']]), ('s4', [['s0', 's2']])])],
1: [OrderedDict([('s5', [['s4']])]), OrderedDict([('s4', [['s0', 's2']])])],
0: [OrderedDict([('s5', [['s3']])]), OrderedDict([('s3', [['s0', 's1']])])]}
Graphviz is necessary to obtain these visualizations
visualization_path(model,
path=paths[0],
target_node='s5',
type_analysis='production',
filename='path_0.png')
# Visualization is saved to local directory with the filename path0.png
visualization_path(model,
path=paths[1],
target_node='s5',
type_analysis='production',
filename='path_1.png')
# Visualization is saved to local directory with the filename path1.png
visualization_path(model,
path=paths[2],
target_node='s5',
type_analysis='production',
filename='path_2.png')
# Visualization is saved to local directory with the filename path2.png