Metabolomics Society Forum

I am beginner of studying how to manipulate and process the LC-MS data.
I am curious how can I manually apply XCMS to access LC-MS data in processing of feature detection, retention time correction, peak alignment (grouping), peak filling just like workflow as XCMS Online. To learn the manipulation and processing of the LC-MS data, I apply the data from Saghatelian et al. which consists of 6 knock-out and 6 wild-type samples. I upload and process these data through XCMS Online. The corresponding log shows as below:

General parameters

polarity     positive
    retention time format     minutes
2. Feature detection

method : centWave
    ppm     30
    snthr     6
    peakwidth     10 60
    mzdiff     0.01
    prefilter peaks     3
    prefilter intensity     100
    Feature detection results :

wt15.CDF    [1_WT]     -->  3515  Features.
   wt16.CDF    [1_WT]     -->  3499  Features.
   wt18.CDF    [1_WT]     -->  3033  Features.
   wt19.CDF    [1_WT]     -->  3047  Features.
   wt21.CDF    [1_WT]     -->  2850  Features.
   wt22.CDF    [1_WT]     -->  3235  Features.
   ko15.CDF    [2_KO]     -->  3516  Features.
   ko16.CDF    [2_KO]     -->  3553  Features.
   ko18.CDF    [2_KO]     -->  3097  Features.
   ko19.CDF    [2_KO]     -->  3087  Features.
   ko21.CDF    [2_KO]     -->  3109  Features.
   ko22.CDF    [2_KO]     -->  3201  Features.
3. Retention time correction

method : obiwarp
    profStep     0.1
4. Alignment

method : peakgroups
    bw     5
    mzwid     0.025
    minfrac     0.5
    minsamp     1
5. FillPeaks
6. Diffreport

class1     1_WT
    class2     2_KO
    statistics.threshold.pvalue     0.01
    statistics.diffReport.value     into
7. Additional Plots & Statistics

I try to use XCMS instruction xcmsSet to prepare analyte data from these LC-MS data jsut like as following.

library(xcms)
PATH <- "ms_netcdf/"
files <- list.files(PATH, pattern="*.CDF", recursive=TRUE, full.names=TRUE)
xset <- xcmsSet(files)

The problem is

According to previous code, total number of peaks is 4721 generated by

xcmsSet constructor which is not consist with the result shown in previous log file. I guess that XCMS automatically construct the xcmsSet object. The procedure maybe contain several individual steps such as xcmsRaw, retcor, group. How can I reproduce the same results provided by XCMS Online?[/list]
I had also tried to detect feature by using the following instructions

wt15 <- xcmsRaw("ms_netcdf/WT/wt15.CDF");
wt15.cw.peakList <- wt15.wt.peakList <- findPeaks.centWave(wt15, ppm=30, snthresh=6, peakwidth=c(10, 60), mzdiff=0.01, prefilter=c(3, 100))

The number of peaks detected is also 3515 which is consist with XCMS Online log file. So I repeat the previous code for data loading and peak detecting of each sample. The following problem is how to do retention time correction? According to help of XCMS document, I found instruction of retcor.obiwarp. However, this instruction belongs to xcmsSet-class method. How can I perform the following retention time correction? This really confused me. Please give me your advising.
Thanks for all.

Best regards,
Hsiang

Hope this help:

files <- list.files(path=myPath, pattern="xml", full.names = TRUE, recursive = TRUE)
xset <- xcmsSet(files, method="centWave", polarity="positive", prefilter=c(3,100), ppm=30, peakwidth=c(10,60), snthr=10)
xsetR <- retcor(xset, method="obiwarp", profStep=0.1, plottype="deviation")
xsetR <- group(xsetR, bw=5, mzwid=0.025, minfrac=0.5, minsamp=1)
xset.finale <- fillPeaks(xsetR)

I aloso suggest to read THIS (http://http://www.bioconductor.org/packages/2.9/bioc/vignettes/xcms/inst/doc/xcmsPreprocess.pdf)
Best

Dear Recca,
Thanks for your suggestion. I had tried above codes line by line. However, while running the second instruction, there was an error message.
Is there any configuration modified? Or I neglect something else?
The version and platform of my R:

R version 2.13.1 (2011-07-08)
Platform: i386-pc-mingw32/i386 (32-bit)
OS: Windows 7

Best regards,
Hsiang

Here I will report some of my examination.

The following R code is now successfully running.

files <- list.files(FILEPATH, full.names=TRUE, recursive=TRUE)
xset <- xcmsSet(files, method="centWave", prefilter=c(3,100), ppm=30, peakwidth=c(10,60), snthresh=6)

Comparing with Ricca's code, some parameters was modified in my solution.

1. Attribute of polarity should be removed.
2. "snthr=10" should be modified as "snthresh=6"

Now, the step of xcmsSet had been worked.  :)

However, I found that the detected number of features (peaks) for each sample is inconsistent with, i.e. a little more than, that produced by XCMS Online. I guess that the number of detected features in the log file provided by XCMS Online had been trimmed due to the retention time correction and peak alignment. But I am not sure.

Besides, while adding attribute of polarity to xcmsSet, the error message I mentioned before - error function (classes, fdef, mtable): unable to find an inherited method for function "findPeaks", for signature "NULL", will be shown. Actually, the polarity is an argument of xcmsSet, not of findPeaks.centWave. Therefore, I reasonably guess that was caused by  inheritance problem of R package.

xcmsSet(files = NULL, snames = NULL, sclass = NULL, phenoData = NULL,
        profmethod = "bin", profparam = list(),
        polarity = NULL, lockMassFreq=FALSE, start=0,
      mslevel=NULL, nSlaves=0, progressCallback=NULL,...)

As previous guess, I still can not confirm my thoughts, because I am not really familiar with R language. If someone could solve my questions, I really appreciate your response.

Best regards,
Hsiang

It's a bug and reproducible with the latest xcms version. Will be fixed in the next release.

I still see this error:

Loading required package: methods
ecoli1_600_pos_1: Error in function (classes, fdef, mtable)  : 
  unable to find an inherited method for function "findPeaks", for signature "NULL"
Calls: xcmsSet -> findPeaks -> <Anonymous>
Execution halted

for code:

xSet = xcmsSet(inputFile, NULL, NULL, NULL, "bin", list(), "positive", 1);

Is it normal ? :)

P.S. Info:

> sessionInfo()
R version 2.11.1 (2010-05-31) 
x86_64-pc-linux-gnu 

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8    
 [5] LC_MONETARY=C              LC_MESSAGES=en_US.UTF-8   
 [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       

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

other attached packages:
[1] CAMERA_1.4.2 xcms_1.22.1 

loaded via a namespace (and not attached):
[1] graph_1.26.0 RBGL_1.24.0  tools_2.11.1