checkBatchEffects: Identify remaining batch effects in expression data
In holgerman/HT12ProcessoR: Preprocessing Illuminas HT12 v4 data

Identification of transcripts where association with batches is still stronger than expected by chance after batch-adjustment. Stronger than chance (i.e. overinflation) is defined as association stronger p-value after Bonferroni correction. All defined Batches will be tested, i.e. 'Sentrix.Barcode', 'processingbatch', and 'strangebatch' in the table from slot '$chipsamples' if more than one category of the batch is found. Probes are not filtered but annotated when found to be overinflated. This function makes use of package MatrixEQTL in order to allow fast ANOVA on matrices. Therefore, some messages from these functions report doing an "eQTL analysis" allthough in fact an ANOVA is done on batch effects

checkBatchEffects(
  ht12object,
  paramfile = NULL,
  showPlots = T,
  round4ANOVAcheck = 5
)

`ht12object`	A list object of class HT12prepro created with function removeBatchEffects()
`paramfile`	Path to the file specifying parameters
`round4ANOVAcheck`	parameter for control how identical Results of MATRIXEQTL and standard ANova have to be. Provided number rounds the F statistic value. Applied in function runMAtrixEQTLAnova called by calcAnovaSentrixRunSpecialbatchViaMatrixEQTL2() . Note that slightly different results may occur when e.g. using two completely nested covariates

A list object of class HT12prepro where the slot with probe-related attributes of the current processing-stage named '$genesdetail' is updated as well as the slot with the history of the commands named '$history'. QQ plots of the association are shown.

holgerman/HT12ProcessoR documentation built on June 5, 2021, 9:18 a.m.