mitoODE {mitoODE} | R Documentation |
The mitoODE
package implements the differential model fitting used
in the “Dynamical modelling of phenotypes in a genome-wide RNAi
live-cell imaging assay” paper and includes the code to reproduce the presented figures and results.
See the vignette for details about the model.
getp0(y, p0=NULL, sd=0) fitmodel(y, p0, pconst, nfits=1, sd=0, mc.cores=1, best=TRUE)
y |
A matrix with 4 columns containing the cell count time series, such
as the one returned by |
p0 |
A vector of 10 numeric parameters, such as returned by |
pconst |
A vector of 10 numeric parameters, containing |
nfits |
The number of fits to be done. Default is 1. |
sd |
A numeric indicating the standard deviation of some Gaussian noise added to the initial conditions before fitting. Default is 0. |
mc.cores |
A numeric indicating how many parallel cores should be used to fit the cell count time series. Default is 1. |
best |
A logical. See return value. |
The function fitmodel
fits the cell count data to the differential equation model described in the vignette, using
the penalized Levenberg-Marquardt nls.lm
algorithm.
getp0
returns a vector of 10 parameters that can be used as initial conditions.
If TRUE
, fitmodel
returns the best fitted parameters, otherwise returns the all n
fitted parameters.
Gregoire Pau, pau.gregoire@gene.com, 2012
loadFittedData
, figure1
, plotfit
## read spot spotid <- 1000 y <- readspot(spotid) ## fit data pconst <- c(g.kim=0.025, g.kmi=0.57, g.mit0=0.05, p.lambda=4) p0 <- getp0() pp <- fitmodel(y, p0, pconst) round(pp, 2) ## plot fitted data plotfit(spotid, pp)