control_metrics: bead array control processing/QC: get metrics from control...

In ttriche/sesamizeGEO: the poor man's methylation suite (cf. `sesamizer`)

bead array control processing/QC: get metrics from control probe signals

Description

mostly lifted from Sean Maden's recountmethylation supplement https://github.com/metamaden/recountmethylationManuscriptSupplement modified slightly to take advantage of rgSet annotations and to Go Big, i.e. this works reasonably well on fairly large rgSets (like all of GEO)

Usage

control_metrics(rgSet, dft = NULL, baseline = 3000, biotin.baseline = 1)

Arguments

rgSet	an rgSet with control probe intensities and annotation
dft	optional single-row data.frame of thresholds (default)
baseline	offset for Extension green based background (3000)
biotin.baseline	baseline offset for biotin staining probes (1)

Value

           a matrix (cols = metrics, rows = samples)

See Also

          flag_control_failures

Examples

  
if (exists("MPAL_rgSet")) {

  # TARGET MPAL data, from Alexander et al, Nature 2018
  cm <- control_metrics(MPAL_rgSet)

  library(ComplexHeatmap)
  Heatmap(t(cm), name="metric", column_names_side="top",
          column_names_gp=gpar(fontsize=6), row_names_side="left")

  library(circlize)
  flagcols <- colorRamp2(c(0, 1), c("white", "darkred"))
  flagged <- flag_control_failures(cm, platform="epic")
  all_samples_passed <- ifelse(colSums(flagged) == 0, 1, 0)
  all_probes_passed <- ifelse(rowSums(flagged) == 0, 1, 0)
  Heatmap(t(flagged), name="failed", col=flagcols, column_names_side="top",
          column_split=all_probes_passed, column_names_gp=gpar(fontsize=6), 
          row_split=all_samples_passed, row_names_side="left")

} 

ttriche/sesamizeGEO documentation built on Nov. 12, 2023, 5:42 p.m.