DTA.dynamic.estimate: Estimation of synthesis and decay rates upon perturbation

Description

DTA.dynamic.estimate uses an experiment, given by a phenotype matrix, data matrix and the number of uridines for each gene to estimate synthesis and decay rate of the genes.

Usage

DTA.dynamic.estimate(phenomat = NULL,datamat = NULL,tnumber = NULL,ccl = NULL,mRNAs = NULL,reliable = NULL,mediancenter = TRUE,usefractions = "LandT",LtoTratio = NULL,ratiomethod = "tls",largest = 5,weighted = TRUE,relevant = NULL,check = TRUE,error = TRUE,samplesize = 1000,confidence.range = c(0.025,0.975),bicor = TRUE,condition = "",upper = 700,lower = 500,save.plots = FALSE,resolution = 1,folder = NULL,fileformat = "jpeg",totaloverwt = 1,sr.vs.dr.folds.lims = c(-5,5),te.vs.to.folds.lims = c(-6,6),robust = FALSE,clusters = "sr",ranktime = NULL,upperquant = 0.8,lowerquant = 0.6,notinR = FALSE,RStudio = FALSE,simulation = FALSE,sim.object = NULL)

Arguments

phenomat

A phenotype matrix, containing the design of the experiment as produced by DTA.phenomat. Columns are name, fraction (U=unlabebeld, L=labeled, T=total), time and nr (=replicate number). Rows represent individual experiments.

datamat

A matrix, containing the measurements from U, L and T, according to the design given in phenomat. Matrix should only contain the rows of phenomat as columns.

tnumber

Integer vector, containing the numbers of uridines. Elements should have the rownames of datamat.

ccl

The cell cycle length of the cells.

mRNAs

Estimated number of mRNAs in a cell (optional).

reliable

Vector of 'reliable' genes, which are used for parameter estimation.

mediancenter

Should the quotient Labeled/Total resp. Unlabeled/Total be rescaled to a common median over it's replicates before building the genewise median.

usefractions

From which fractions should the decay rate be calculated: "LandT", "UandT" or "both".

LtoTratio

Coefficient to rescale Labeled/Total. Is estimated from the data, if not specified. See ratiomethod.

ratiomethod

Choose the regression method to be used, possible methods are: "tls", "bias" and "lm". For details, see supplemental material of Sun et al. (see references).

largest

Percentage of largest residues from the first regression not to be used in the second regression step. For details, see supplemental material of Sun et al. (see references).

weighted

Should the regression be weighted with 1/(Total^2 + median(Total))?

relevant

Choose the arrays to be used for halflives calculation, vector due to nr (=replicate number) in phenomat.

check

If check = TRUE, control messages and plots will be generated.

error

If TRUE, the measurement error is assessed by means of an error model and resampling to gain confidence regions.

samplesize

Error model samplesize for resampling.

confidence.range

Confidence region for error model as quantiles. Interval should be between 0 and 1.

bicor

Should the labeling bias be corrected?

condition

String, to be added to the plotnames.

upper

Upper bound for labeling bias estimation. For details, see supplemental material of Sun et al. (see references).

lower

Lower bound for labeling bias estimation. For details, see supplemental material of Sun et al. (see references).

save.plots

If save.plots = TRUE, control plots will be saved.

resolution

Resolution scaling factor for plotting.

folder

Path to the folder, where to save the plots.

fileformat

Fileformat for plots to be saved. See plotit function (LSD package).

totaloverwt

Will be available in the very near future for comparative DTA data.

sr.vs.dr.folds.lims

Limits of the folds plot (dr vs sr).

te.vs.to.folds.lims

Limits of the folds plot (LT vs LE).

robust

If robust = TRUE, LE resp. LT is chosen instead of sr resp. dr.

clusters

should the dr vs sr folds be plotted with clusters, choose 'sr', 'dr' for cluster selection or 'none' to omit it

ranktime

at which time should the rankgain be calculated, default is the last column

upperquant

upper quantile for cluster selection

lowerquant

lower quantile for cluster selection

notinR

Should plots be not plotted in R.

RStudio

For RStudio users. Suppresses the opening of a new device, as RStudio allows only one.

simulation

True, if data was generated by DTA.generate.

sim.object

Simulation object created by DTA.generate.

Value

triples: Mapping of each fraction and experiment to its corresponding column in the data matrix.
plabel: The labeling efficiency. For details, see the vignette.
LtoTratio: Estimated ratio of labeled to total fraction.
UtoTratio: Estimated ratio of unlabeled to total fraction.
LtoUratio: Estimated ratio of labeled to unlabeled fraction.
correcteddatamat: Labeling bias corrected data matrix.
drmat: Decay rates for each replicate. The last column gives the median decay rates.
dr: Median decay rates. The last column of drmat.
dr.confidence: Confidence regions of decay rates.
hlmat: Half-lives for each replicate. The last column gives the median half-lifes.
hl: Median half-lives. The last column of hlmat.
hl.confidence: Confidence regions of half-lives.
TEmat: Total expression for each replicate. The last column gives the median total expression values.
TE: Median total expression values. The last column of TEmat.
TE.confidence: Confidence regions of total expression values.
LEmat: Labeled expression for each replicate. The last column gives the median labeled expression values.
LE: Median labeled expression values. The last column of LEmat.
LE.confidence: Confidence regions of labeled expression values.
UEmat: Unlabeled expression for each replicate. The last column gives the median unlabeled expression values. (Only if unlabeled values exist in the experiment)
UE: Median unlabeled expression values. The last column of UEmat. (Only if unlabeled values exist in the experiment)
UE.confidence: Confidence regions of unlabeled expression values.
srmat: Synthesis rates for each replicate. The last column gives the median synthesis rates.
sr: Median synthesis rates. The last column of srmat.
sr.confidence: Confidence regions of synthesis rates.
LtoTmat: Labeled to total ratio for each replicate. The last column gives the median labeled to total ratios.
LtoT: Median labeled to total ratios. The last column of LtoTmat.
LtoT.confidence: Confidence regions of labeled to total ratios.
UtoTmat: Unlabeled to total ratio for each replicate. The last column gives the median unlabeled to total ratios.
UtoT: Median unlabeled to total ratios. The last column of UtoTmat.
UtoT.confidence: Confidence regions of unlabeled to total ratios.
Rsrmat: Rescaled synthesis rates for each replicate, if parameter mRNAs is specified. The last column gives the median synthesis rates.
Rsr: Rescaled median synthesis rates. The last column of Rsrmat.
globaldrmat: Decay rate for each replicate. Reciprocally weighted by the total expression. Last element contains (weighted) median decay rate.
globaldr: (Weighted) median decay rate.

References

C. Miller, B. Schwalb, K. Maier, D. Schulz, S. Duemcke, B. Zacher, A. Mayer, J. Sydow, L. Marcinowski, L. Doelken, D. E. Martin, A. Tresch, and P. Cramer. Dynamic transcriptome analysis measures rates of mRNA synthesis and decay in yeast. Mol Syst Biol, 7:458, 2011. M. Sun, B. Schwalb, D. Schulz, N. Pirkl, L. Lariviere, K. Maier, A. Tresch, P. Cramer. Mutual feedback between mRNA synthesis and degradation buffers transcript levels in a eukaryote. Under review. B. Schwalb, B. Zacher, S. Duemcke, D. Martin, P. Cramer, A. Tresch. Measurement of genome-wide RNA synthesis and decay rates with Dynamic Transcriptome Analysis (DTA/cDTA). Bioinformatics.

Examples

Run this code

dataPath = system.file("data", package="DTA")
load(file.path(dataPath, "Miller2011dynamic.RData"))

### for control plots set 'check = TRUE' ###

res = DTA.dynamic.estimate(Sc.phenomat.dynamic,Sc.datamat.dynamic,Sc.tnumber,ccl = 150,mRNAs = 60000,reliable = Sc.reliable.dynamic,LtoTratio = rep(0.1,7),check = FALSE)

Run the code above in your browser using DataLab