0
19 months ago by
bhgyu30
bhgyu30 wrote:

Hi guys,

I can't seem to find an answer to this issue. It has been flagged before but without a successful solution.

cwp <- CentWaveParam(peakwidth=c(5,20), ppm=15, snthresh=5, prefilter = c(0,0))
xod <- findChromPeaks(raw, param = cwp)​
xod <- adjustRtime(xod, param = ObiwarpParam())
Sample number 6 used as center sample.
Aligning SJ45.CDF against SJ50.CDF ...
Error: BiocParallel errors
element index: 1, 2, 3, 4, 5, 6, ...
first error: Dimensions of profile matrices of files SJ50.CDF and SJ45.CDF do not match!

I also tried running adjustRtime on the raw data files, but without success.

Strangely enough, when I run the old XCMS pipeline with rector instead of adjustRtime it all works, but I'd really like to try the new functions!

Any insights?

R version 3.4.0 (2017-04-21)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows >= 8 x64 (build 9200)


modified 19 months ago by Johannes Rainer1.4k • written 19 months ago by bhgyu30
1
19 months ago by
Johannes Rainer1.4k
Italy
Johannes Rainer1.4k wrote:

Hi,

this has to do most likely with the binning of the data. Try using a different binSize for the ObiwarpParam, e.g. binSize = 2. And what version of xcms are you using? Could you please provide the full output of the sessionInfo?

Would it be possible for you to share the data so I could have a closer look what's happening (eventually a smaller set)?

cheers, jo

1

Hi Johannes,

Here's the full sessionInfo.

Sadly I won't be able to share the data - not my data and 12GB.

I did try various binSizes, but that didn't work :-( Could it be somewhere in the initial upload steps? Maybe it has to be binned before being RT corrected?

R version 3.4.0 (2017-04-21)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows >= 8 x64 (build 9200)

Matrix products: default

locale:
[1] LC_COLLATE=English_United States.1252  LC_CTYPE=English_United States.1252    LC_MONETARY=English_United States.1252
[4] LC_NUMERIC=C                           LC_TIME=English_United States.1252

attached base packages:
[1] parallel  stats     graphics  grDevices utils     datasets  methods   base

other attached packages:
[1] RColorBrewer_1.1-2  xcms_1.52.0         MSnbase_2.2.0       ProtGenerics_1.8.0  mzR_2.10.0
[6] Rcpp_0.12.13        BiocParallel_1.10.1 Biobase_2.36.2      BiocGenerics_0.22.1

loaded via a namespace (and not attached):
[1] git2r_0.19.0           compiler_3.4.0         BiocInstaller_1.26.1   plyr_1.8.4             iterators_1.0.8
[6] tools_3.4.0            zlibbioc_1.22.0        MALDIquant_1.16.4      digest_0.6.12          memoise_1.1.0
[11] preprocessCore_1.38.1  tibble_1.3.4           gtable_0.2.0           lattice_0.20-35        rlang_0.1.2
[16] Matrix_1.2-9           foreach_1.4.3          curl_3.0               knitr_1.17             httr_1.3.1
[21] withr_2.0.0            devtools_1.13.3        S4Vectors_0.14.7       IRanges_2.10.5         stats4_3.4.0
[26] multtest_2.32.0        grid_3.4.0             impute_1.50.1          R6_2.2.2               snow_0.4-2
[31] survival_2.41-3        XML_3.98-1.9           RANN_2.5.1             limma_3.32.10          ggplot2_2.2.1
[36] MASS_7.3-47            splines_3.4.0          scales_0.5.0           pcaMethods_1.68.0      codetools_0.2-15
[41] MassSpecWavelet_1.42.0 mzID_1.14.0            colorspace_1.3-2       affy_1.54.0            lazyeval_0.2.0
[46] munsell_0.4.3          doParallel_1.0.11      vsn_3.44.0             affyio_1.46.0 

1

Good news is from the version you are using (1.52.0) to the current version (2.99.10) there have been many changes, improvements and fixes. Unfortunately we couldn't merge all fixes into the Bioconductor 3.5 release version because of some nasty git/svn complications.

Now, Bioconductor release 3.6 is on its way (release next week) with the updated and (presumably fixed) xcms. In the meantime you could try to install the current Bioconductor developmental version (3.6); should be save now, since it will be released soon:

library(BiocInstaller)
useDevel()
biocLite(c("mzR", "MSnbase", "xcms"))

Thank you! Will give this a shot and report back.

1

Ooo lala it worked :-)



Thanks for your help! Oh and if you have any recommendations for settings adjusted for Q-TOF for these new functions, let me know. I'm using as much as I can from my settings for the older xcmsSets, but the new workflow is quite a bit different and there's not much info out there as to adjustments. For example, how do I extract the results? Diffreport obviously won't work with the new expression sets.

1

The settings and parameters are essentially the same as in the old interface - just a little revamped to fit todays Bioconductor (and reuse functionality and classes from other packages). Eventually have a look at the new xcms vignette in the updated package - I try to explain some additional stuff and diagnostic plots there too.