Bond-Calculus and Continuous Pi

This repository contains two biologically-inspired process calculi, the bond-calculus and the continuous pi-calculus.

Quickstart

The simplest way of trying these languages is via the prebuilt docker image thomasdwright/bondwb. You can download the image and run cpiwb using the command,

xhost +local:root && docker run -it -v $(pwd):/host -w /host -v /tmp/.X11-unix:/tmp/.X11-unix:rw -e DISPLAY thomasdwright/bondwb cpiwb

or run bondwb using the command,

xhost +local:root && docker run -it -v $(pwd):/host -w /host -v /tmp/.X11-unix:/tmp/.X11-unix:rw -e DISPLAY thomasdwright/bondwb bondwb

These commands only depend on having a working install of Docker. They also do a little additional setup:

• The current directory will be mounted at /host within the image, allowing you to use any models within the current directory.
• This enables X11-forwarding between your host and the image, allowing you to view the plots windows produced by the tool.

See Usage for some examples of how to get started with the tools.

Building

If you wish to compile/edit the source code yourself, you can download the source code from this repo and compile the tools using stack.

The languages are configured to build using stack in a isolated environment within another Docker image -- thomasdwright/bond-env -- containing all of its dependancies.

This depends only on a working installation of Docker, stack, and git. This is also only known to work on a Linux or OSX (whilst Docker has Windows support, stack's docker integration and graphical output via X11 do not: stack issue #2421).

Stack can be installed via the command (see https://docs.haskellstack.org/en/stable/README/):

wget -qO- https://get.haskellstack.org/ | sh

whilst installation instructions for docker are available at (https://docs.docker.com/install/).

Then you can clone the repository using,

git clone https://github.com/twright/bondwb.git

Next enter the directory and setup stack,

cd bondwb
stack setup

One can build the library and all of its dependencies using the command

stack build

Running

Once the build is finished the bond-calculus' command line interface, bondwb, can be run via the command:

stack exec bondwb

whilst continuous pi's commandline interface, cpiwb, can be run via the command

stack exec cpiwb

If graphical plot output is required, it can be displayed outside of the container using X11-forwarding, however requires giving the machine access to your display by first running the command:

xhost +local:root

Usage

Using cpiwb

Once you have started cpiwb, you should see the following prompt

Welcome to the Continuous Pi-calculus Workbench (CPiWB).
Type "help" for help.

CPiWB:> 

Next we will load an example model stored in the file models/enzyme.cpi:

species E(e,s)
  = e.0;
species S(e,s)
  = s.(ES(e,s));
species ES(e,s)
  = tau<1>.(E(e,s)|P(e,s));
species P(e,s) = tau<0>.0;

process Pi
   = [100] S(e,s) || [10] E(e,s) || [0] P(e,s) || [0] ES(e,s)
  :{ e-s @ 1 };

using the following command:

CPiWB:> load models/enzyme.cpi
Loading: models/enzyme.cpi
Done. Type "env" to view.

It is now possible to simulate the resulting model (via generating ODEs) and plot the result using the command,

CPiWB:> plot Pi 0 10 100

This specifies that we should plot the concentration of each species in the process Pi from timepoints 0 to 10 using 100 steps, and results in the display of the following plot window:

It is also possible to use the command help to list all available commands.

Using bondwb

Once you have started bondwb, you should see the following prompt,

Welcome to the Biological Continuous Pi-calculus Workbench (BioWB).
Type "help" for help.

BioWB:> 

Next we will load an example model stored in the file models/enzymemm.biocpi:

species S = s -> P;

species P = 0;

species E = e -> E;

kinetic law MichaelisMenten(k, K; C, S) = k*C*S/(K + S);

affinity network M(v,k) = {
  e, s at rate MichaelisMenten(v, k);
}

process Pi = [100] S || [10] E with network M(1.1, 2.0);

using the following command:

BioWB:> load models/enzymemm.biocpi
Loading: models/enzymemm.biocpi
Done. Type "env" to view.

It is now possible to simulate the resulting model (via generating ODEs) and plot the result using the command,

BioWB:> plot Pi 0 10 100

resulting in the following plot window:

We can also perform a stochastic simulation of the model using StochPy's implementation of Gillespie's Stochastic Simulation Algorithm:

BioWB:> stochpy Pi 0.01 10 "tauleap"

This specifies a simulation of process Pi with the continuous concentrations of the models scaled to discrete variables by a factor 0.5, simulated for 10 units of time using the tauleap method (this can be any method supported by StochPy), resulting in the following plot:

It is also possible to use the command help to list all available commands.

Updating docker images

The build environment image is described by Dockerfile; if needed it can be rebuilt via the command:

docker build -t thomasdwright/bondwb-env -f Dockerfile .

whilst the deployment image can be rebuilt via the command:

stack image container