modelBF: Gene regulation with positive feedback with basal...

In CharlotteJana/pdmppoly: Moment Approximation for Polynomial PDMPs

Description

This PDMP models a gene regulation mechanism similar to genePolyF, where we have one gene and a positive feedback loop. The difference is that in both discrete states transcription takes place, but with different rates a_0, a_1, where a_0 < a_1. Transcription and translation are considered as one step and are not modeled separately. In PROM, this model is referred to as Model BF+, therefore it is named genePdmpBF and genePolyBF here.

Usage

genePdmpBF

genePolyBF

Format

genePdmpBF is an object of class pdmpModel,
genePolyBF is an object of class polyPdmpModel.

Simulation

The simulations in PROM were done with slot times set to

• from = 0, to = 1000, by = 0.1.

The following parameter sets were simulated:

• k01 = 0.02, k10 = 0.02, a = 1, b = 0.2

• k01 = 0.02, k10 = 0.02, a = 7, b = 0.2

Source

The parameter values do not rely on real data.

Examples

library(spray)
#------ code to generate the pdmpModel version -----

genePdmpBF <- new("pdmpModel",
                  descr = "Model BF: positive feedback with basal transcription",
                  parms = list(b = 0.5, a0 = 1, a1 = 3, k10 = 1, k01 = 0.5), 
                  init = c(f = 1, d = 1),
                  discStates = list(d = 0:1),
                  dynfunc = function(t, x, parms) {
                    df <- with(as.list(c(x, parms)), {
                      switch(d+1, a0 - b*f, a1 - b*f)
                    })
                    return(c(df, 0))
                  }, 
                  ratefunc = function(t, x, parms) {
                    return(with(as.list(c(x, parms)), switch(d + 1, k01*f, k10)))
                  }, 
                  jumpfunc = function(t, x, parms, jtype) {
                    c(x[1], 1 - x[2])
                  }, 
                  times = c(from = 0, to = 100, by = 0.1), 
                  solver = "lsodar")

#------ code to generate the polyPdmpModel version -----

genePolyBF <- new("polyPdmpModel",
                  descr = "Model BF: pos. feedback with basal transcription (polynomial version)",
                  parms = list(b = 0.5, a0 = 1, a1 = 3, k10 = 1, k01 = 0.5), 
                  init = c(f = 1, d = 1), 
                  discStates = list(d = 0:1),
                  dynpolys = quote(list(
                    list(overall = -b*lone(1,2),
                         specific = list(a0, a1))
                  )),
                  ratepolys = quote(list(  
                    list(k01*lone(1,2), k10)
                  )),
                  jumpfunc = function(t, x, parms, jtype) {
                    c(x[1], 1 - x[2])
                  }, 
                  times = c(from = 0, to = 100, by = 0.1), 
                  solver = "lsodar")

#------- comparison of the models --------------

identical(sim(genePdmpBF, outSlot = FALSE, seed = 20),
          sim(genePolyBF, outSlot = FALSE, seed = 20))

CharlotteJana/pdmppoly documentation built on Sept. 4, 2019, 4:40 p.m.