GenoGAM-class: GenoGAM class
In gstricker/fastGenoGAM: A GAM based framework for analysis of ChIP-Seq data

Description Usage Arguments Value Methods (by generic) Slots Author(s) Examples

This is the class the holds the complete model as well all hyperparameters and settings that were used to fit it. It extends the RangedSummarizedExperiment class by adding a couple of more slots to hold hyperparameters and settings. The 'assays' slot holds the basepair fit and standard deviation. Additionally all knot positions and beta coefficients will be stored in the 'smooths' slot in order to be able to make use of the piecewise function that produces the fit. For information on the slots inherited from SummarizedExperiment check the respective class.

The GenoGAM constructor, not designed to be actually used, by the user. Rather to be a point of reference and documentation for slots and how to access them.

GenoGAM(..., ggd = NULL, fromHDF5 = FALSE, split = FALSE)

## S4 method for signature 'GenoGAM'
design(object)

## S4 method for signature 'GenoGAM'
sizeFactors(object)

## S4 method for signature 'GenoGAM'
getSettings(object)

## S4 method for signature 'GenoGAM'
getFamily(object)

## S4 method for signature 'GenoGAM'
colData(x)

## S4 method for signature 'GenoGAM'
getParams(object)

## S4 method for signature 'GenoGAM'
getCoefs(object)

## S4 method for signature 'GenoGAM'
getKnots(object)

## S4 method for signature 'GenoGAM,missing'
assay(x, i)

## S4 method for signature 'GenoGAM'
fits(object)

## S4 method for signature 'GenoGAM'
se(object)

## S4 method for signature 'GenoGAM'
pvalue(object)

## S4 method for signature 'GenoGAM'
colnames(x)

## S4 method for signature 'GenoGAM'
dimnames(x)

## S4 method for signature 'GenoGAM'
is.HDF5(object)

## S4 method for signature 'GenoGAM,GRanges,ANY,ANY'
x[i]

`...`	Slots of the GenoGAM class. See the slot description.
`ggd`	The initial GenoGAMDataSet object. Only needed if fromHDF5 is TRUE.
`fromHDF5`	A convenience argument to create a GenoGAM object from the already computed fits that are stored as HDF5 files
`split`	A logical argument indicating if the model was fitted in a per-chromosome fashion or not. Only needed if fromHDF5 is TRUE.
`object, x`	For use of S4 methods. The GenoGAM object.
`i`	A GRanges object (only for subsetting)

An object of the type GenoGAM.

design: An accessor to the design slot
sizeFactors: An accessor to the sizeFactors slot
getSettings: An accessor to the settings slot
getFamily: An accessor to the family slot
colData: An accessor to the factorialDesign slot.
getParams: An accessor to the params slot
getCoefs: An accessor to the coefs slot
getKnots: An accessor to the knots slot
assay: The accessor to the fits and standard errors
fits: An accessor to the fits
se: An accessor to the standard errors
pvalue: An accessor to the pvalues
colnames: column names of GenoGAM
dimnames: The names of the dimensions of GenoGAM
is.HDF5: A boolean function that is true if object uses HDF5 backend
[: Additional subsetting by single brackets

family: A GenoGAMFamily object
design: The formula of the model
sizeFactors: The offset used in the model.
factorialDesign: The factorial design used. The same as colData in the GenoGAMDataSet
params: All hyperparameters used to fit the data. The parameters estimated by cross validation can also be found here. But the parameters used in cross validation are in the settings slot.
settings: A GenoGAMSettings object representing the global settings that were used to compute the model.
coefs: The coefficients of the knots
knots: The relative knot positions
hdf5: A logical slot indicating if the data is stored in a HDF5 format on hard drive

Georg Stricker georg.stricker@in.tum.de

## creating test GenoGAM object
gg <- makeTestGenoGAM()
gg

## using accessors
design(gg)
sizeFactors(gg)
getSettings(gg)
getFamily(gg)
colData(gg)
getParams(gg)
getCoefs(gg)
getKnots(gg)
rowRanges(gg)
assay(gg)
assays(gg)
fits(gg) 
se(gg)