Skip to main content

Show Posts

This section allows you to view all Show Posts made by this member. Note that you can only see Show Posts made in areas you currently have access to.

Messages - Jan Stanstrup

R / Re: Find samples with high deviation in retentiontime, XCMS
I did this a few days ago for a colleague exactly the way that you describe.
Replace `xcms_p_POS_g_r_g` with your object.

Code: [Select]

investigate_data <- tibble(rtime =        rtime(xcms_p_POS_g_r_g, adjusted = FALSE, bySample = TRUE),
                          adjustedRtime = rtime(xcms_p_POS_g_r_g, adjusted = TRUE, bySample = TRUE),
                          file =          basename(fileNames(xcms_p_POS_g_r_g))
                          ) %>%
    unnest(c(rtime, adjustedRtime)) %>%
    mutate(diff =          adjustedRtime - rtime)

investigate_data %>%
    filter(abs(adjustedRtime-250)<1) %>% # <-- focus on a small RT range where the outlier is clear
    group_by(file) %>%
    arrange(diff) %>% # <-- use desc if looking for a positive difference
    slice(1) %>%
    ungroup %>%
    arrange(diff)     # <-- use desc if looking for a positive difference

Courses and training / [COURSE] Introduction to Nutritional Metabolomics
Course dates
06 July 2020 - 10 July 2020

Place: Copenhagen

Info, sign-up, program

The course will provide an overview of LC-MS based untargeted metabolomics and its application in nutrition. It will be delivered using a mixture of lectures, hands-on data preparation and analysis, computer-based practical sessions, and discussions. Visits to wet labs and instructions on human sample preparation procedures is included but with minimal hands-on.

The students will go through common steps in a typical metabolomics study using a real-life case. This case study includes collected plasma (or urine) samples from a nutritional intervention. The sample preparation and analysis on UPLC-QTOF has been conducted and the students will further process and analyse the acquired data with various freeware tools (e.g. R, XCMS, MZmine and Metaboanalyst). They will finally work on identification of relevant metabolites using several web-based structure elucidation tools. The course will conclude by presentations of reports generated by the students based on the case study.

The course will be structured as initial short lectures on theory followed by hands-on exercises, which will teach the students to transfer the theoretical information to practice.
Other / Re: Software for generating deisotoped and charge state deconvoluted MS1 peaks
Deconvolution in the sense of one feature/variable per peak doesn't exist out of the box in any software for untargeted LC-MS that I am aware of.

The most common tools for untargeted data would be XCMS+CAMERA or MZmine. There is nothing straight forward about doing this on untargeted data though.
Mass spectrometry / Re: Experimental design for multiple-batch study
If you have "paired samples" where you are not interested in the differences within the pairs it can reduce variance to put them in the same batch. For example if you have several persons you might consider having the same person in the same batch for all time course samples.

But as you say. Spread evenly conditions and randomize within the batch.

Use plenty of QC samples so that you afterwards can correct for potential intra- and inter-batch drifts.
R / Re: XCMS in R
Quadrupole instruments sometimes save data as chromatograms and not spectra.
You can try converting your original files with msconvert from Proteowizard and add: --simAsSpectra --srmAsSpectra
R / Re: Looking for help: Dimensions of profile matrices do not match !
It shows that indeed grouping did make groups. So I guess retcor must mean that non of those groups adhere to the missing and extra requirements. Try increasing those. You have a lot of samples. So you are requiring that for a peak to be used for the correction model all samples, except 1, has found the peak. And that groups where, in any 1 sample, more than one peak was merged in a peak group, cannot be used either.

try something like this as a starting point:

Code: [Select]
missing  = 200
extra   = 100

If you are including very sparse samples, like artificial mixtures or blanks, you might need to go even higher.
span =1 is also very strict but should not affect the error.
XCMS / Re: Consequences of using Centwave for profile mode data
I would ask for here complete settings at this point and see if you can replicate at all. the prefilter setting in particular might explain how she was able to run the analysis at all. The grouping step might also have put humpty dumpty somewhat back together again but you should see a huge npeaks number in the peak table. a positive "mzdiff" I guess also might join things up to look sensible on the surface.