Subsetting a XcmsExperiment object after alignment does not revert retention times of detected chrom peaks

[jorainer]

Calling e.g. xmse[4, keepAdjustedRtime = FALSE] on an XcmsExperiment object after retention time alignment will correctly revert the retention times, but not revert the "rt", "rtmin" and "rtmax" values of chromatographic peaks. The retention time information for chromatographic peaks will still be the adjusted retention times. Note that this does only affect the XcmsExperiment, but not the XCMSnExp result object.

This is a clear bug we need to fix ASAP in both the release and devel branch!

A reproducible example:

library(xcms)
library(MsExperiment)
library(testthat)
fls <- c(system.file('cdf/KO/ko15.CDF', package = "faahKO"),
                    system.file('cdf/KO/ko16.CDF', package = "faahKO"),
                    system.file('cdf/KO/ko18.CDF', package = "faahKO"))
df <- data.frame(mzML_file = basename(fls),
                 dataOrigin = fls,
                 sample = c("ko15", "ko16", "ko18"))
mse <- readMsExperiment(spectraFiles = fls, sampleData = df)
cwp <- CentWaveParam(noise = 10000, snthresh = 40, prefilter = c(3, 10000))
xmse <- findChromPeaks(mse, param = cwp)
pdp <- PeakDensityParam(sampleGroups = rep(1, 3))
xmse <- groupChromPeaks(xmse, param = pdp, add = FALSE)
xmse <- adjustRtime(xmse, PeakGroupsParam(span = 0.4))

## Subsetting with/without keeping adjusted retention times
a <- xmse[3, keepAdjustedRtime = TRUE]
b <- tmp[3, keepAdjustedRtime = FALSE]
    
## retention times are expected to be different
expect_false(all(rtime(a) == rtime(b)))

## retention times of chrom peaks would also be expected to be different
## but they are not
expect_false(all(chromPeaks(a)[, "rt"] == chromPeaks(b)[, "rt"]))
Error: all(chromPeaks(a)[, "rt"] == chromPeaks(b)[, "rt"]) is not FALSE

`actual`:   TRUE 
`expected`: FALSE

[jorainer]

In addition to this, we are also not consistent between the filterFile() method, that has parameter keepAdjustedRtime = TRUE - and the [,XcmsExperiment that has keepAdjustedRtime = FALSE. So, we should, in addition, also change the default for this parameter!