Launch the modeling process
Launch the modeling process with parameters set with
This method model the synthesis, degradation and processign rates after their estimation by the constructor function
newINSPEcT. Estimated rates are not guaranteed to optimally describes provided input data yet.
To this purpose, modeled rates can be generated and genes can be assigned to a transcriptional regulatory mechanism.
Modeled rates can be accessed via the method
viewModelRates and gene classification according
to the regulatory mechanism can be accessed by
geneClass. The modeling procedure can be set by the
user by modyging the parameters via
1 2 3 4 5
An object of class INSPEcT
A numeric, indicatindg the seed to be set for reproducible results
Parallelization parameters for bplapply. By default bpparam()
Either NULL or logical. If logical indicates whether to output some text
during computation or not, if NULL it takes the information from the object
When modeling many genes, parallelization is strongly suggested to reduce computational time. Since all genes run independently, the computational time is diveded by the number of cores used/available. However, when modeling more than 500 genes, it may happen that a single gene returns an error that escapes the try/catch controls of INSPEcT. With the parallel mode, the error will propagate on all genes that have been computed with the same processor (or core). To avoid this, the computation could be splitted in chunks and the whole data set can be obtaied by combining the chunks (see Examples).
An object of class INSPEcT with modeled rates
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
data('rpkms', package='INSPEcT') tpts <- c(0, 1/6, 1/3, 1/2, 1, 2, 4, 8, 16) tL <- 1/6 mycerIds <- newINSPEcT(tpts, tL, rpkms$foursu_exons, rpkms$total_exons, rpkms$foursu_introns, rpkms$total_introns, BPPARAM=SerialParam()) mycerIdsOneGene <- mycerIds ## View modeling parameters modelingParams(mycerIdsOneGene) ## Run the modeling in a reproducible way (setting seed) mycerIdsOneGene <- modelRates(mycerIdsOneGene, seed=1, BPPARAM=SerialParam()) ## view modeled synthesis rates viewModelRates(mycerIdsOneGene, 'synthesis') ## view gene classes geneClass(mycerIdsOneGene) ## Divide a parallel computation into chunks ## Not run: nCores(mycerIds) <- parallel::detectCores() chunkSize <- 100 splitIdx <- ceiling(c(1:nGenes(mycerIds))/chunkSize) chunks <- lapply(split(mycerIds, splitIdx), modelRates) mycerIdsModeled <- do.call('combine', chunks) ## End(Not run)
Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.