Conversation
- Fix issue #812 : parameters `minFraction` and `maxFeatures` were not correctly considered in `plotChromPeakDensity()`.
- Add parameter `columns` to `chromPeakData()` for `XCMSnExp` and `XcmsExperiment` classes to allow extraction of selected columns.
- Add support for parameter `columns` to the `chromPeakData()` method for `XcmsExperimentHdf5` objects.
- Parameter `skipFilled` is now supported for `featureSpectra,XcmsExperiment`.
- Add unit tests checking that parameter `skipFilled` works correctly for `featureSpectra()` for `XcmsExperimentHdf5` objects.
|
@sneumann , sorry to be a bit pushy here, but we should try to manage getting this PR into xcms before the next Bioconductor release. This contains some important bug fixes. |
- `featureArea()` and `ChromPeakAreaParam()` gain additional parameter `minMzWidthPpm` that allows to define a minimal guaranteed m/z width of the reported feature areas respectively regions to integrate data for gap filling (issue #756).
|
Note: added in the meantime also an additional parameter |
R/DataClasses.R
Outdated
| ) | ||
| mzmax = "function", | ||
| minMzWidthPpm = "numeric"), | ||
| contains = "Param",, |
There was a problem hiding this comment.
what is the intention behind ,, here ?
There was a problem hiding this comment.
Oh, wow! does not make any sense, true. I'll remove. Thanks!
R/XcmsExperiment.R
Outdated
| mtch <- as.matrix( | ||
| findMatches(sps$chrom_peak_id, | ||
| rownames(.chromPeaks(object))[pindex])) | ||
| rownames(xcms:::.chromPeaks(object))[pindex])) |
There was a problem hiding this comment.
good catch - was a leftover from some tests. I'll remove.
| ## rtdiff = diff(range(fd$rtmed[z]))) | ||
| ## }) |> do.call(what = rbind) | ||
| ## | ||
| ## Thoughts on how to combine overlapping features: |
There was a problem hiding this comment.
so within a file overlapping features would result from non-optimal feature detection.
Across samples overlapping features could come from non-optimal correspondence analysis.
There was a problem hiding this comment.
yes - we've had now a dataset where we've seen some strange overlap in features. the idea of this function would be sort of quality check of the correspondence settings. but I'm not yet sure whether that's helpful. so, I just jotted down the idea to not forget to look more into it.
There was a problem hiding this comment.
I like the idea of having a corner of QC methods being calculated, with later the option to export those as mzQC values. but indeed, not just yet.
2 participants
In this PR:
XcmsExperimentobjects toXCMSnExp(issue Convert XcmsExperiment to XCMSnExp object error #807).columnstochromPeakData()to allow extraction of selected columns.skipFilledtofeatureSpectra()forXcmsExperimentandXcmsExperimentHdf5to avoid extraction of spectra for gap-filled chromatographic peaks.spectraData()method from Spectra.