processData: Processing expression time series

In ahonkela/tigre: Transcription factor Inference through Gaussian process Reconstruction of Expression

Description

processData further processes time series data preprocessed by puma or lumi.

processRawData performs similar processing for other data.

Both functions return ExpressionTimeSeries objects that can be used as input for the functions GPLearn and GPRankTargets.

Usage

  processData(data, times = NULL, experiments = NULL,
    do.normalisation = TRUE)
  processRawData(rawData, times, experiments = NULL,
    is.logged = TRUE, do.normalisation = ifelse(is.logged, TRUE, FALSE))

Arguments

data	The preprocessed data from mmgMOS in the puma package (an exprReslt object) or from the lumi package (a LumiBatch object).
rawData	Raw data matrix to be used. Each row corresponds to a gene and each column to a data point.
times	Observation times of each data point. If unspecified or NULL, processData attempts to infer this from phenoData(data) field containing 'time' in the name.
experiments	The replicate structure of the data indicating which expression data points arise from which experiments. This should be an array in integers from 1 to N with length equal to the number of data points. By default all the data points are assumed to be from same replicate.
is.logged	Indicates whether the expression values are on log scale or not. Normalisation of non-logged data is unsupported.
do.normalisation	Indicates whether to perform the normalisation.

Details

The expression data (and percentiles, if available) are normalized by equalising the mean of log-expression in each time points. In processData, a normal distribution is then fitted into the data with distfit.

Value

An ExpressionTimeSeries object containing all provided information.

Author(s)

Antti Honkela, Jonatan Ropponen

See Also

GPLearn, GPRankTargets.

Examples

  ## Load a mmgmos preprocessed fragment of the Drosophila developmental
  ## time series
  data(drosophila_mmgmos_fragment)

  ## Process the data (3 experiments containing 12 time points each)
  drosophila_gpsim_fragment <- processData(drosophila_mmgmos_fragment,
    experiments=rep(1:3, each=12))

ahonkela/tigre documentation built on Aug. 6, 2021, 12:08 p.m.