FWSEplot: Plot reported vs. empirical quality.
In ReQON: Recalibrating Quality Of Nucleotides

Description Usage Arguments Details Value Author(s) Examples

View source: R/FWSEplot.R

Plots reported vs. empirical quality scores. Also calculates and outputs Frequency-Weighted Squared Error (FWSE) and reports FWSE on the plot. The points are shaded according to the relative frequency of bases assigned that quality score.

1
2
3

FWSEplot(ErrRates, QualFreq, FWSE_out = TRUE, col = "blue", max_freq = 0.25, 
   lim = c(0, length(QualFreq) - 1), collegend = TRUE, xlabel = "Reported Quality", 
   ylabel = "Empirical Quality", main_title = "Reported vs. Empirical Quality")

`ErrRates`	vector of empirical error rates on the Phred scale.
`QualFreq`	vector of relative frequencies of quality scores.
`FWSE_out`	option to output FWSE and report FWSE on the plot. Default = TRUE.
`col`	color of plotted points. Currently allows for only "blue" and "red". Default = "blue".
`max_freq`	Maximum value for color scale, where all quality scores with relative frequency greater than this value are shaded the darkest color. Because the colors are evenly distributed between 0 and max_freq, changing this maximum value may change the color of many points. Default = 0.25.
`lim`	common axis limits for both the x-axis and y-axis. Default = c(0, length(QualFreq) - 1).
`collegend`	include color legend. Default = TRUE.
`xlabel`	x-axis label. Default = "Reported Quality".
`ylabel`	y-axis label. Default = "Empirical Quality".
`main_title`	title. Default = "Reported vs. Empirical Quality".

FWSEplot plots the reported quality score against the empirical quality score and reports FWSE. If the quality scores accurately reflect the probability of a sequencing error, then the points should fall close to the 45-degree line and FWSE should be close to zero. If the input vectors are $ErrRatesBefore and $QualFreqBefore from the ReQON output, this function will create the bottom left diagnostic plot that is output from ReQON. Similarly, if the input vectors are $ErrRatesAfter and $QualFreqAfter from ReQON output, then the bottom right diagnostic plot is created.

For more details and interpretation, see the vignette by: browseVignettes("ReQON").

In addition to the plot, FWSEplot calculates and outputs Frequency-Weighted Squared Error (FWSE), a measure of how close the points lie to the 45-degree line.

Christopher Cabanski ccabansk@genome.wustl.edu

## Create relative frequency example data
require( stats )
after <- dnorm( c( 0:40 ), mean = 30, sd = 8 )
  after <- after / sum( after )
err_rate <- c( 0:40 ) + rnorm( 41, mean = 0, sd = 5) 
  err_rate[ which( err_rate < 0 ) ] <- (-1) * err_rate[ which( err_rate < 0 ) ]
  ## to guarantee that all values are positive

## plot and calculate FWSE
FWSEplot( err_rate, after, col = "red" )