Calculate correlation between technical replicates

Calculate and return list of correlation between paired technical replicates in a dataset

Usage

calcReplicateCorr(
  dgeObject,
  group = NULL,
  corrThreshold = 0.8,
  return = "all",
  method = "pearson",
  ignoreZero = FALSE
)

Arguments

dgeObject: DGEList object with barcode counts with technical replicates.
group: Column name in sample metadata with replicate information (string). Correlation can only be calculated between pairs of replicates.
corrThreshold: Threshold distinguishing good vs bad correlation between technical replicates (decimal). Default = `0.8`.
return: Which values to return, one of `good`, `bad`, `all` (string). Default = `all`.
method: Method for correlation, one of `spearman`, `pearson` (string). Default = `pearson`.
ignoreZero: Remove barcodes where both replicates have 0 counts before calculating correlation (boolean). Including zeros can increase the spearman correlation, especially for samples with few barcodes detected. Default = `FALSE`.

Value

Returns a plot of the read counts per barcode (row) in a data frame

Examples

data(test.dge)
calcReplicateCorr(test.dge, group = "group")
#> 10_High_dose   11_Vehicle   12_Vehicle  13_Low_dose  14_Low_dose  15_Low_dose 
#>    0.9998576    0.9988335    0.9992809    0.9999569    0.9999779    0.9998940 
#>  16_Low_dose 17_High_dose   18_Vehicle  1_High_dose  2_High_dose  3_High_dose 
#>    0.9998965    0.9998269    0.9995630    0.9990497    0.9975044    0.9997354 
#>    4_Vehicle   5_Low_dose  6_High_dose    7_Vehicle    8_Vehicle   9_Low_dose 
#>    0.9975635    0.9999389    0.9994988    0.9997089    0.9992883    0.9997059 
#>           T0 
#>    0.9983848