SPEM: S-system parameter estimation method
In SPEM: S-system parameter estimation method

Description Usage Arguments Details Value Methods Author(s) Examples

This function calculates parameters of S-system from entire time series matrix.

1 2	## S4 method for signature 'ExpressionSet' SPEM(TS_eSet, n = 3, sparsity = 0.2, lbH = -3, ubH = 3, lbB = 0, ubB = 10)

`TS_eSet`	Time series data in ExpressionSet class. assayData: Matrix with n metabolite in row and m time points in column. phenoData: phenoData type. The sample data.frame should include the label "time", which represents the values of time points.
`n`	Positive integer, SPEM will guess initial beta n times.
`sparsity`	A positive number. In order to force the interaction matrix to be sparse, interactions with absolute value smaller than "sparsity" will be set to zero.
`lbH`	Lower boundary value of h.
`ubH`	Upper boundary value of h.
`lbB`	Lower boundary value of beta.
`ubB`	Upper boundary value of beta.

In this SPEM package, we aim to reconstruct gene networks from time-series expression data using the S-system model. The input dataset should be as an ExpressionSet data container, describing, in assayData, expression data for n genes (rows) and m time points (columns), along with a vector of length m, which records the exact values of time points, thus showing the sample intervals in phenoData. SPEM will calculate the parameters alpha, G, beta and H of the S-system function set that best fits the dataset.

`alpha, G, beta, H`	Parameters of the reconstructed S-system.
`IniBeta`	Guess of the IniBeta value (Picked randomly by SPEM itself).
`error`	Regression error.

signature(TS_eSet = "ExpressionSet"): This method is created for function SPEM.

Yang, X-Y., Dent, Jennifer E. and Nardini, C.

#########Generate Toy Model #######
#########
# If you want to calculate SOS dataset in this package, please read our vignette###
#Real dataset takes a long time to calculate. You may want to try function 'row_optimize' to compute it in parallel###


toy_expression_data<-matrix(data=abs(rnorm(12)),nrow=3,ncol=4, dimnames=list(paste("G",c(1:3),sep=''), paste("tp",c(0,2,4,6),sep="_")))
toy_timepoints_data<-data.frame(index=c(1:4), label=paste("tp",c(0,2,4,6),sep='_'), time=c(0,2,4,6),row.names=paste("tp",c(0,2,4,6),sep='_'))
toy_varMetadata<-data.frame(labelDescription=c("Index number","Label Detail", "Time points values"),row.names=c("index","label","time"))
toy_phenoData<-new("AnnotatedDataFrame", data=toy_timepoints_data,varMetadata=toy_varMetadata)
toy_ExpressionSet<-new("ExpressionSet", exprs=toy_expression_data,phenoData=toy_phenoData)

#########Set parameters #######
n<- 1 
sparsity<- 0.2
lbH<- -3
ubH<- 3
lbB<- 0
ubB<- 10
#########Calculate results #######

result<-SPEM(toy_ExpressionSet,n,sparsity,lbH,ubH,lbB,ubB)