lmmsDE-methods: Differential expression analysis using linear mixed effect...

In lmms: Linear Mixed Effect Model Splines for Modelling and Analysis of Time Course Data

Description

Function to fit a linear mixed effect model splines to perform differential expression analysis. The lmmsDE function fits LMM models with either a cubic, p-spline or cubic p-spline basis and compares the models to the null models. The type of basis to use is specified with the basis argument.

Usage

lmmsDE(data, time, sampleID, group, type,
experiment, basis, knots,keepModels, numCores)

Arguments

data	data.frame or matrix containing the samples as rows and features as columns
time	numeric vector containing the sample time point information.
sampleID	character, numeric or factor vector containing information about the unique identity of each sample
group	character, numeric or factor vector containing information about the group (or class) of each sample
type	character indicating what type of analysis is to be performed. Options are "time" to identify differential expression over time, "group" to identify profiles with different baseline levels (intercepts), and "time*group" an interaction between these two . Use "all" to calculate all three types.
experiment	character describing the experiment performed for correlation handling. Use "all" for data-driven selection of model; "timecourse" for replicated experiments with less variation in individual expression values (e.g. model organism, cell culture), "longitudinal1" for different intercepts and "longitudinal2" for different intercepts and slopes.
basis	character string. What type of basis to use, matching one of "cubic" smoothing spline as defined by Verbyla et al. 1999, "p-spline" Durban et al. 2005 or a "cubic p-spline".
knots	can take an integer value corresponding to the number of knots for the chosen basis or by default calculated as in Ruppert 2002. Not in use for the 'cubic' smoothing spline basis.
keepModels	alternative logical value if you want to keep the model output. Default value is FALSE
numCores	alternative numeric value indicating the number of CPU cores to be used for parallelization. Default value is automatically estimated.

Details

lmmsDE extends the LMMS modelling framework to permit tests between groups, across time, and for interactions between the two implemented as described in Straube et al. 2015.

Value

lmmsDE returns an object of class lmmsde containing the following components:

DE	data.frame returning p-values and adjusted p-values using Benjamini-Hochberg correction for multiple testing of the differential expression testing over time, group or their interaction.
modelsUsed	numeric vector indicating the model used to fit the data. 1=linear mixed effect model spline (LMMS) with defined basis ('cubic' by default) 2 = LMMS taking subject-specific random intercept, 3 = LMMS with subject specific intercept and slope.
predTime	data.frame containing predicted values based on linear model object or linear mixed effect model object.
predGroup	data.frame containing predicted values based on linear model object or linear mixed effect model object.
predTime	data.frame containing predicted values based on linear model object or linear mixed effect model object.
predTimeGroup	data.frame containing predicted for the time*group model values based on linear model object or linear mixed effect model object.
modelTime	a list of class lme, containing the models for every feature modelling the time effect.
modelGroup	a list of class lme, containing the models for every feature modelling group effect.
modelTimeGroup	a list of class lme, containing the models for every feature modelling time and group interaction effect.
type	an object of class character, describing the test performed either time, group, time*group or all.
experiment	an object of class character describing the model used to perform differential expression analysis.

References

Durban, M., Harezlak, J., Wand, M. P., & Carroll, R. J. (2005). Simple fitting of subject-specific curves for longitudinal data. Stat. Med., 24(8), 1153-67.

Ruppert, D. (2002). Selecting the number of knots for penalized splines. J. Comp. Graph. Stat. 11, 735-757

Verbyla, A. P., Cullis, B. R., & Kenward, M. G. (1999). The analysis of designed experiments and longitudinal data by using smoothing splines. Appl.Statist, 18(3), 269-311.

Straube J., Gorse A.-D., Huang B.E., & Le Cao K.-A. (2015). A linear mixed model spline framework for analyzing time course 'omics' data PLOSONE, 10(8), e0134540.

See Also

summary.lmmsde, plot.lmmsde

Examples

## Not run: 
data(kidneySimTimeGroup)
lmmsDEtest <-lmmsDE(data=kidneySimTimeGroup$data,time=kidneySimTimeGroup$time,
              sampleID=kidneySimTimeGroup$sampleID,group=kidneySimTimeGroup$group)
summary(lmmsDEtest)
## End(Not run)

lmms documentation built on May 2, 2019, 6:53 a.m.