A python package that allows users to integrate expression data into Copasi (.cps) models. Additionally, it allows users to use copasi model and perform the following tasks: steadystate calculation, metabolic control analysis, and parameter scan.
The pycoexp framework is developed using Python 3.
$ pip install git+https://github.com/surajsept/pycoexp.git
Once you have checked the steps above. You may consider refering to the Beginners Guide. Or, just start a python console to:
import pycoexp.tasks
task = pycoexp.tasks.tasks()
# integrate expression values and save updated copasi models in folder 'updatedModels'
task.integrate_expression(filepath_CPSmodel='model.cps', filepath_expdata='ExpData.csv',
filepath_mapping='mapping.csv', foldername='updatedModels/')
Name | Model_ID | ... |
---|---|---|
Name | Exp1 | Exp2 | ... |
---|---|---|---|
here, "Name" corresponds to the gene name/symbol, "Model_ID" stands for the ID in SBML model and "Exp1..." stand for experiment/observation name.
Additionally,
# get steady state
Concentration, Flux = task.steadystate(filepath_CPSmodel='model.cps')
# perform metabolic control analysis
ControlCoefficients = task.mca(filepath_CPSmodel='model.cps', system_variable='concentration', verbose=True)
# do parameter scan
Conc, Fluxes = task.scan(filepath_CPSmodel='model.cps', parameter_name='NAMPT', E_T_or_k1='E_T', lb=0.1, ub=1.0, n=10, rescaling=True)
# do time-course simulation
Conc, Fluxes = task.time_course(filepath_CPSmodel='model.cps', duration=100, stepsize=0.1)
This is a development version, hosted as a private git repository, and people interested in contributing can request access by contacting Suraj Sharma (surajsept@gmail.com).