Error: No peak groups found after using centWave method

September 24, 2013, 01:47:33 PM

My LC/MS data were generated using Agilent Q-TOF 6520, which is considered to be high-resolution.

First, I just used the universal code for peak finding, as shown below.

Code: [Select]

xset<- xcmsSet(mzXMLfiles)

I could group the peaks, correct the retention time, fill the missing peaks and eventually get final result visualization.

However, when I tried to use the centWave method for those high-resolution data, I could do group, but not go through the retcor step. Please refer to the following codes and error/warning message.

Code: [Select]

> xset<- xcmsSet(mzXMLfiles, method="centWave", snthresh=10, ppm=25, peakwidth=c(20,50), prefilter=c(3, 100), mzCenterFun="wMean", integrate=1, mzdiff=-0.001, fitgauss=FALSE)
> xset=group(xset)
> xset2 <- retcor(xset,  family = "symmetric", plottype = "mdevden")

Quote

Retention Time Correction Groups: 1
Error in .local(object, ...) :
Not enough ``well behaved'' peak groups even for linear smoothing of retention times
In addition: Warning message:
In .local(object, ...) : Too few peak groups, reverting to linear method

I did play with the arguments such as snthresh, peakwidth, prefilter, fitgauss, unfortunately I failed whatever I tried. The error message is:

Quote

Error in .local(object, ...) :
No peak groups found for retention time correction

In addition, I tried to use obiwarp method for retention time correction after failing to use the default loess method (as shown above). Unfortunately, my desktop was down always whenever I ran the following code.

Code: [Select]

> xset2 <- retcor(xset,  method= "obiwarp", plottype = "deviation")

Any suggestion for my above-mentioned issues? Thanks you so much in advance.

Best wishes,

Jeff

Re: Error: No peak groups found after using centWave method

Reply #1 – September 30, 2013, 03:05:42 AM

Jeff,

A couple of things:
1) The data that you have used for the centWave was centroided data?
2) How many peaks do you have in 'xset' ?
3) How many groups do you have in 'xset' ?

I would suspect that this is an issue of parameter settings. You might want to turn on the sleep parameter to help and see what is going on. Also try the group.neareest method.

Paul

Re: Error: No peak groups found after using centWave method

Reply #2 – January 15, 2014, 03:01:04 AM

hello
i am a new user of XCMS and i have similar problem. i want to use CentWave and i have the same error message. i use centroid data but in R, i have warnings message like what my data are in profile mode
initially my files are .RAW, i convert them to mzXML with Proteowizard but i have always the message that my files are in profile mode
bellow a part of my script

xset1 <- xcmsSet(method="centWave", ppm=10 ,peakwidth=c(6,30), noise= 1000000, mzdiff=0.01, prefilter=c(4,1000000), snthresh = 6,fitgauss=TRUE)
can you help me please?
binta