Welcome to netplotlib’s documentation!¶

Netplotlib is an extension to NetworkX and matplotlib to draw and analyze reaction network diagrams in SBML or Antimony strings with ease. Netplotlib supports visualization of quantities such as flux and species rate of change. Netplotlib provides functions to visualize network model ensemble.

Installation¶

Netplotlib is available through PyPI. To install, run:

$ pip install netplotlib

Netplotlib is also available by default in Tellurium.

Quick Start¶

To start using netplotlib, install the package using pip or use Tellurium.

This section demonstrates how to load in a model and draw network diagrams. To start with, lets import netplotlib package and define a model. Here, we use a simple feed forward loop as an example.

import netplotlib as npl

AntimonyStr = '''
$Xi -> S1; k0*Xi
S1 -> S2; k1*S1
S2 -> S3; k2*S2
S1 -> S3; k3*S1
S3 -> $Xo; k4*S3

Xi = 3; Xo = 2
k0 = 0.46; k1 = 0.73; k2 = 0.64;
k3 = 0.51; k4 = 0.22
'''

Next, create an Network object.

net = npl.Network(AntimonyStr)

To generate network diagrams, simply run draw().

net.draw()

Tutorial¶

Network Configuration¶

Network object supports following properties for configuring the network diagram.

scale: scaling factor for layout algorithm
fontsize: fontsize for labels
edgelw: linewidth of edges
nodeColor: node color
reactionNodeColor: reaction node color
labelColor: label color
labelReactionIds: boolean flag for labeling reaction ids
reactionColor: edge color
modifierColor: modifier edge color
boundaryColor: boundary node color
nodeEdgeColor: node edge color
nodeEdgelw: linewidth of node edges
highlight: list of species ids or reaction ids to highlight
hlNodeColor: node color of highlighted nodes
hlNodeEdgeColor: node edge color of highlighted nodes
drawReactionNode: boolean flag for drawing reaction nodes
breakBoundary: boolean flag for breaking all boundary species into separate nodes
tightLayout: boolean flag for tighter node layout
analyzeFlux: boolean flag for visualizing flux
analyzeRates: boolean flag for visualizing species rate of changes
analyzeColorHigh: color to use for higher values during analysis
analyzeColorLow: color to use for lower values during analysis
analyzeColorMap: colormap to use for analysis.
analyzeColorScale: boolean flag for using colormaps. Setting this true ignore analyzeColorHigh and analyzeColorLow
drawInlineTimeCourse: boolean flag for plotting inline time-cource simulation
simTime: value for simulation duration
forceAnalysisAtSimTime: boolean flag for running analysis at the end of simTime
plotColorbar: boolean flag for visualizing color bar
inlineTimeCourseSelections: list of species to plot for time-course simulation
customAxis: use custom matplotlib.pyplot.axis object for the diagram
layoutAlgorithm: specify layout algorithm

You can define these properties and run draw() to generate customized network diagrams.

Detached Boundary¶

Detach common boundary species by setting breakBoundary=True.

net = npl.Network(npl.testmodels.REPRESSILATOR)
net.draw()
net.breakBoundary = True
net.draw()

Flux Visualization¶

Visualize flux using colormap via setting analyzeFlux=True. Set analyzeColorScale=True to scale the colormap to minimum and maximum values. You can supply your own colormaps. To plot colorbar, set plotColorbar=True.

net.analyzeColorScale = True
net.analyzeFlux = True
net.analyzeColorMap = 'viridis'
net.plotColorbar = True

Rate of Changes Visualization¶

Visualize species rates of change using colormap via setting analyzeRates=True. Set analyzeColorScale=True to scale the colormap to minimum and maximum values. The resulting plot will show species rate of change at t=simTime. You can supply your own colormaps. To plot colorbar, set plotColorbar=True.

net.analyzeColorScale = True
net.analyzeRates = True
net.analyzeColorMap = 'viridis'
net.simTime = 3000
net.plotColorbar = True

Inline Time-course Plot¶

Plot inline time-course plot along with network diagrams with matching colors by setting drwaInlineTimeCourse=True. The simulation runs for t=simTime. You can selection specific species to plot by passing a list to inlineTimeCourseSelections property.

net.drawInlineTimeCourse = True
net.inlineTimeCourseSelections = ['MKKK_P', 'MKKK', 'MAPK_PP']
net.simTime = 4000

Highlighting¶

Highlight specific nodes by passing a list to highlight property. Choose the highlight colors using hlNodeColor and hlNodeEdgeColor properties.

net.drawReactionNode = False
net.nodeEdgelw = 3
net.highlight = ['S1']
net.draw()

Selecting Layout Algorithm¶

Currently, netplotlib supports following layout algorithms:

‘kamada-kawai’

‘spring’

‘dot’

‘neato’

‘twopi’

By default, netplotlib uses Kamada-Kawai layout algorithm. Certain layout algorithms require external graphviz to be configured and pygraphviz package installed.

NetworkEnsemble Configuration¶

NetworkEnsemble object supports following properties for configuring the network diagram.

scale: scaling factor for layout algorithm
fontsize: fontsize for labels
edgelw: linewidth of edges
nodeColor: node color
reactionNodeColor: reaction node color
labelColor: label color
labelReactionIds: boolean flag for labeling reaction ids
reactionColor: edge color
modifierColor: modifier edge color
boundaryColor: boundary node color
nodeEdgeColor: node edge color
nodeEdgelw: linewidth of node edges
highlight: list of species ids or reaction ids to highlight
hlNodeColor: node color of highlighted nodes
hlNodeEdgeColor: node edge color of highlighted nodes
edgeLabel: boolean flag for displaying edge weights
edgeLabelFontSize: fontsize of edge weight labels
drawReactionNode: flag for drawing reaction nodes
breakBoundary: boolean flag for breaking all boundary species into separate nodes
weights: list of custom weights to override
edgeTransparency: boolean flag for changing the transparency of the edges accordin to edge weights
plottingThreshold: value of threshold to prevent from displaying weighted edges
removeBelowThreshold: boolean flag for preventing weighted edges below plottingThreshold from displaying
analyzeFlux: boolean flag for visualizing flux
customAxis: use custom matplotlib.pyplot.axis object for the diagram
layoutAlgorithm: specify layout algorithm

Grid Plot¶

Plot a grid plot of network diagrams of individual models in the list by running drawNetworkGrid() function.

net.drawNetworkGrid(3, 3)

Weighted Network Diagram¶

Combine models in the list and generate a one network diagram where the edges are weighted according to the frequency. To generate a weighted network diagram, run drawWeightedDiagram() function. Below are images of an example where an ensemble of three models are combined into a weighted network diagram.

net.drawWeightedDiagram()

It is possible to set a threshold where edges below the threshold are removed from the resulting network diagram. To set a threshold, use plottingThreshold and removeBelowThreshold properties. For example, if plottingThreshold=0.5 and removeBelowThreshold=True, any edges that appear in less than half of the model ensemble will be ignored. However, sometimes you might want to put the ignored reactions back while keeping the layout. To do so, set removeBelowThreshold=False while keeping the plottingThreshold.

net.removeBelowThreshold = True
net.plottingThreshold = 0.5
net.drawWeightedDiagram()
net.removeBelowThreshold = False
net.drawWeightedDiagram()

Test Cases¶

Netplotlib comes with set of test cases. All test cases are available under netplotlib.testmodels submodule.

Examples¶

Aspartate Metabolism¶

Feedback Loop¶

MAPK Cascade¶

Glycolysis¶

Network Ensemble¶

netplotlib API¶

Single Network¶

netplotlib.Network(model)¶

Creates a new Network object.

Parameters:	model – SBML or Antimony string of a model

An instance of Network object accepts following properties:

scale: scaling factor for layout algorithm

fontsize: fontsize for labels

edgelw: linewidth of edges

nodeColor: node color

reactionNodeColor: reaction node color

labelColor: label color

labelReactionIds: boolean flag for labeling reaction ids

reactionColor: edge color

modifierColor: modifier edge color

boundaryColor: boundary node color

nodeEdgeColor: node edge color

nodeEdgelw: linewidth of node edges

highlight: list of species ids or reaction ids to highlight

hlNodeColor: node color of highlighted nodes

hlNodeEdgeColor: node edge color of highlighted nodes

drawReactionNode: boolean flag for drawing reaction nodes

breakBoundary: boolean flag for breaking all boundary species into separate nodes

tightLayout: boolean flag for tighter node layout

analyzeFlux: boolean flag for visualizing flux

analyzeRates: boolean flag for visualizing species rate of changes

analyzeColorHigh: color to use for higher values during analysis

analyzeColorLow: color to use for lower values during analysis

analyzeColorMap: colormap to use for analysis.

analyzeColorScale: boolean flag for using colormaps. Setting this true ignore analyzeColorHigh and analyzeColorLow

drawInlineTimeCourse: boolean flag for plotting inline time-cource simulation

simTime: value for simulation duration

forceAnalysisAtSimTime: boolean flag for running analysis at the end of simTime

plotColorbar: boolean flag for visualizing color bar

inlineTimeCourseSelections: list of species to plot for time-course simulation

customAxis: use custom matplotlib.pyplot.axis object for the diagram

layoutAlgorithm: specify layout algorithm

Network.draw(show=True, savePath=None, dpi=150)¶

Draw network diagram

Parameters:	show – flag to show the diagram savePath – path to save the diagram dpi – dpi settings for the diagram

Network.getLayout(returnState=False)¶

Return the layout of the model

Parameters:	returnState – boolean flag for returning the networkx.Graph object
Returns pos:	Dictionary of all nodes and corresponding coordinates

Network.reset()¶: Resets all properties

Network.savefig(path, dpi=150)¶

Save network diagram to specified location

Parameters:	path – path to save the diagram dpi – dpi settings for the diagram

Network.setLayout(pos)¶

Set custom layout and bypass whe is generated by the layout algorothm

Parameters:	pos – Dictionary of all nodes and corresponding coordinates

Network Ensemble¶

netplotlib.NetworkEnsemble(models)¶

Creates a new NetworkEnsemble object.

Parameters:	models – list of SBML or Antimony strings of models

An instance of NetworkEnsemble object accepts following properties:

scale: scaling factor for layout algorithm

fontsize: fontsize for labels

edgelw: linewidth of edges

nodeColor: node color

reactionNodeColor: reaction node color

labelColor: label color

labelReactionIds: boolean flag for labeling reaction ids

reactionColor: edge color

modifierColor: modifier edge color

boundaryColor: boundary node color

nodeEdgeColor: node edge color

nodeEdgelw: linewidth of node edges

highlight: list of species ids or reaction ids to highlight

hlNodeColor: node color of highlighted nodes

hlNodeEdgeColor: node edge color of highlighted nodes

edgeLabel: boolean flag for displaying edge weights

edgeLabelFontSize: fontsize of edge weight labels

drawReactionNode: flag for drawing reaction nodes

breakBoundary: boolean flag for breaking all boundary species into separate nodes

weights: list of custom weights to override

edgeTransparency: boolean flag for changing the transparency of the edges accordin to edge weights

plottingThreshold: value of threshold to prevent from displaying weighted edges

removeBelowThreshold: boolean flag for preventing weighted edges below plottingThreshold from displaying

analyzeFlux: boolean flag for visualizing flux

customAxis: use custom matplotlib.pyplot.axis object for the diagram

layoutAlgorithm: specify layout algorithm

Network.drawWeightedDiagram(show=True, savePath=None, dpi=150)¶

Draw weighted reaction network based on frequency of reactions

Parameters:	show – flag to show the diagram savePath – path to save the diagram dpi – dpi settings for the diagram
Returns allRxn:	list of all reactions in the list of models presented as a pair of reactants and products
Returns count:	normalized count of reactions in allRxn throughout the list of models

Network.drawNetworkGrid(nrows, ncols, auto=False, show=True, savePath=None, dpi=150)¶

Plot a grid of network diagrams

Parameters:	nrows – number of rows ncols – number of columns auto – Automatically configure nrows and ncols based on the number of models. Overrides nrows and ncols. show – flag to show the diagram savePath – path to save the diagram dpi – dpi settings for the diagram

Network.getLayout(returnState=False): Return the layout of the model

Network.reset(): Resets all properties

Network.savefig(path, dpi=150)

Save weighted network diagram to specified location

Parameters:	path – path to save the diagram dpi – dpi settings for the diagram

Network.setLayout(pos)

Set custom layout and bypass whe is generated by the layout algorothm

Parameters:	pos – Dictionary of all nodes and corresponding coordinates

Welcome to netplotlib’s documentation!¶

Installation¶

Quick Start¶

Tutorial¶

Network Configuration¶

Detached Boundary¶

Flux Visualization¶

Rate of Changes Visualization¶

Inline Time-course Plot¶

Highlighting¶

Selecting Layout Algorithm¶

NetworkEnsemble Configuration¶

Grid Plot¶

Weighted Network Diagram¶

Test Cases¶

Examples¶

Aspartate Metabolism¶

Feedback Loop¶

MAPK Cascade¶

Glycolysis¶

Network Ensemble¶

netplotlib API¶

Single Network¶

Network Ensemble¶

Miscellaneous¶

Indices and tables¶