An example of count data for an experiment with spike-ins
pulseRSpikeinsData
A list containing simulated data, model and parameter information.
formulas describe the model for mean read number
counts contain the simulated data
conditions is a data.frame describing sample time point and type
fractions is a character vector to divide samples into groups for
normalisation
formulaIndexes is a list of formula names (see formulas), which are
used in linear combination with coefficients in normFactors to calculate
mean read number
spikeins is a list with two elements:
refGroup is the name of the sample which to use as a reference sample
spikeLists is a list of names of spike-ins which are to be used
for normalisation. It has the same structure as formulaIndexes
allNormFactors is a list of true coefficients which were used for
spike-ins and sample counts normalisations before simulation
par is a list of model parameters
The data set contains simulation of an experiment which measures three
fractions, namely, coded as 'A_fraction', 'B_fraction' and 'C_fraction'.
There are three types of quantities, 'A', 'B' and 'C' for which a kinetic
model is defined, see formulas element in the data set:
A = a, B = a * b ^ time, C = alpha * a * (1 - b ^ time)
where a, b are gene-specific parameters which are unknown, alpha is a parameter which is shared between all genes.
The data are generated for 3 replicates, 3 different time points and for
10 different genes, see elements counts and conditions.
The model considers possibility of cross-contamination with different types
of RNA, which is described by formulaIndexes simply as
formulaIndexes <- list( A_fraction = 'A', B_fraction = c('B', 'C'), C_fraction = c('B', 'C'))
In this case,
the mean read number for a gene is a linear combination of the described
RNA types with weights defined in normFactors.
Spike-ins counts are generated in order to recover normalistion coefficients
which describe how read counts in different samples relate to each other.
Different spike-ins correspond to different types of RNA (e.g. labelled and
unlabelled) and the rule for this relations are defined in the
the spikeins element for this data set.
The true normalisation coefficients which were used for data simulation
are contained in the allNormFactors list.
The true parameter values, which are used for data simulation are in
par element. This also includes the size parameter for the negative
binomial distribution.