Parameter Scan

Copasi supports three types of scan, a regular parameter scan, a repeat scan and sampling from a parametric distributions.

We first build a model to work with throughout the tutorial.

import os
import site
from pycotools3 import model, tasks, viz
import pandas

working_directory = r'/home/ncw135/Documents/pycotools3/docs/source/Tutorials/timecourse_tutorial'
if not os.path.isdir(working_directory):
    os.makedirs(working_directory)

copasi_file = os.path.join(working_directory, 'michaelis_menten.cps')

if os.path.isfile(copasi_file):
    os.remove(copasi_file)

antimony_string = """
model michaelis_menten()
    compartment cell = 1.0
    var E in cell
    var S in cell
    var ES in cell
    var P in cell

    kf = 0.1
    kb = 1
    kcat = 0.3
    E = 75
    S = 1000

    SBindE: S + E => ES; kf*S*E
    ESUnbind: ES => S + E; kb*ES
    ProdForm: ES => P + E; kcat*ES
end
"""

with model.BuildAntimony(copasi_file) as loader:
    mm = loader.load(antimony_string)


mm

The BuildAntimony context manager is deprecated and will be removed in future versions. Please use model.loada instead.

Model(name=michaelis_menten, time_unit=s, volume_unit=l, quantity_unit=mol)

S = tasks.Scan(
    mm, scan_type='scan', subtask='time_course', report_type='time_course',
    report_name = 'ParameterScanOfTimeCourse.txt', variable='S',
    minimum=1, maximum=20, number_of_steps=8, run=True,
)

## Now check parameter scan data exists
os.path.isfile(S.report_name)

True

Two Way Parameter Scan

By default, scan tasks are removed before setting up a new scan. To set up dual scans, set clear_scans to False in a second call to Scan so that the first is not removed prior to adding the second.

## Clear scans for setting up first scan
tasks.Scan(
    mm, scan_type='scan', subtask='time_course', report_type='time_course',
    variable='E', minimum=1, maximum=20, number_of_steps=8, run=False, clear_scan=True,
)

## do not clear tasks when setting up the second
S = tasks.Scan(
    mm, scan_type='scan', subtask='time_course', report_type='time_course',
    report_name = 'TwoWayParameterScanOfTimeCourse.csv', variable='S',
    minimum=1, maximum=20, number_of_steps=8, run=True, clear_scan=False,
)

## check the output exists
os.path.isfile(S.report_name)

True

An arbitrary number of scans can be setup this way. Further, its possible to chain together scans with repeat or random distribution scans.

Repeat Scan Items

Repeat scans are very useful for running multiple parameter estimations and for running stochastic time courses.

## Assume Parameter Estimation task already configured
tasks.Scan(
    mm, scan_type='repeat', subtask='parameter_estimation', report_type='parameter_estimation',
    number_of_steps=6, run=False, ##set run to True to run via CopasiSE
)


## Assume model runs stochastically and time course settings are already configured
tasks.Scan(
    mm, scan_type='repeat', subtask='time_course', report_type='time_course',
    number_of_steps=100, run=False,  ##set run to True to run via CopasiSE
)

<pycotools3.tasks.Scan at 0x1cb859370b8>