simulation: Simulation of a biochemical system with spatial effects

In sbioPN: sbioPN: Simulation of deterministic and stochastic spatial biochemical reaction networks using Petri Nets

Description

These functions simulate a biochemical reacton system with spatial effects parameterized as a Petri Net. sGillespieOptimDirect and sGillespieDirectCR performs pure stochastic simulations, sRungeKuttaDormandPrince45 a pure deterministic integration, sHaseltineRawlings a hybrid of the above. Multiple runs can be performed at once.

See init for a way of defining the model that is close to the way reactions are written.

Usage

## Exact stochastic simulation: 
sGillespieOptimDirect(model, timep, delta=1, runs=1)
sGillespieDirectCR(model, timep, delta=1, runs=1)

## Pure deterministic:
sRungeKuttaDormandPrince45(model, timep, delta=1, ect = 1e-09)

## Hybrid stochastic/deterministic:
sHaseltineRawlings(model, timep, delta=1, runs=1, ect = 1e-09)

Arguments

model	list containing named elements:
timep	It can be either a numeric, indicating for how long (in the same time units as the propensity constants) the process will run, or a functions (R or C), in which case can be used to change the protocol at time intervals. See details.
delta	Interval time at which the state will be saved.
runs	How many runs will be performed.
ect	Precision for the fast reactions.

Details

model is a list containing the following elements:

• model$pre: pre matrix, with as many rows as transitions (reactions), and columns as places (reactants). It has the stoichiometrics of the left sides of the reactions.

• model$post: post matrix, with as many rows as transitions, and columns as places (products). It has the stoichiometrics of the right sides of the reactions.

• model$h: list of propensity constants or functions returning the propensity (with as many elements as transitions).

• model$slow: vector of zeros for slow transitions and ones for fast transitions. Only needed for HaseltineRawlings. Ignored otherwise.

• model$M: initial marking (state) of the system.

• model$place: vector with names of the places.

• model$transition: vector with names of the transitions.

Value

The functions return a list with the following elements:

place	vector with the names of the places if supplied. If not, the function creates names as follows: P1, P2, ...
transition	vector with the names of the transitions if supplied. If not, the function creates names as follows: T1, T2, ...
dt	vector containing the discretized times at which the state is saved (according to delta)
run	list with as many elements as runs. We will describe the first element, run[[1]], as the rest have exactly the same structure. It is also a list, with the following elements:
run[[1]]$M	list with as many elements as places, each of them containing the state of the system sampled according to delta.
run[[1]]$transitions	vector with as many elements as transitions, with the total of time each slow reaction fired.
run[[1]]$tot.transitions	numeric with the summ of run[[1]]$transitions.

See Also

init, atr

Examples

## sbioPN has been tested only on 64 bits machines.
## It may fail in 32 bits architecture.

sbioPN documentation built on May 1, 2019, 8:47 p.m.