Hi, I want to analyze my metabolomic data, but, I have all my raw data in .wiff and .scanwiff files. What is the next step? how can I clean and do the normalization for this data?, how can I do it in R?
Thanks
Adriana
1) Convert with proteowizard's msconvert to mzML
2) Use the XCMS package in R to pre-process data. Probably use CAMERA too for annotation.
3) Use whatever normalization and stats in R is appropriate for your data.
What is better loess regresion o median fold change in LC-MS? Im using XCMS online
Not familiar with XCMS online but loess regression in XCMS I only know from the retention time alignment while fold change would be for the stats.
You'll have to be a bit more specific to get help is my feeling...
Hi, Im using XCMS online, there, option of normalize with the intensity values by either probabilistic quotient or cyclic loess is presented. Im not sure wich choose.
Other option in the same tab is "values to be used for the diffreport. If value="into", integrated peak intensities are used. If value="maxo", maximum peak intensities are used." I dont know if I choose "into" or "maxo" according to the normalization process.
Thanks
Adriana
.
I am not familiar with the normalization in xcms Online. That seems to be specific to their interface.
The other option is about which values to use. "into" is the area under the peaks, maxo is the height of the peaks. So it is not directly related to normalization I would assume. Just they are on the same tab. Normally you'd use "into".
The normalization methods listed are to deal with dilution effects. Some people use these methods to normalize the effect of hydration in plasma/serum, for example. There are certain assumptions associated with those methods, so you should read up on them before you use them. Depending on your sample type, there are other ways to normalize your data (dry/wet weight for tissue, cell count, total signal, etc).
In terms of using area under peaks or max peak height, I think most often you would use area. From my understanding, max peak height is used more for when you have atypical peak shapes ("saw-tooth" peaks). It would be best to have a chat with an analytical chemist about this.
I have four samples in my LC-MS analysis per treatment, and I have 3 treatments. If I see the PCA, score plots shows me there is a biological sample in the first treatment very distant for the rest, however in the others two treatments the distant sample is no the same, how can I know when should I eliminate a biological sample paying attention to the PCA?
Thanks
Adriana
First thing to do is ask yourself if you can explain why a sample is different.
Does the loadings give a hint? -->
If all variables are on one side you have systematic sensitivity/concentration issues.
Does sample analysis order help explain?
Is it a well defined group of variables that cause the difference? What are they? Could they be contaminants?
Would you expect that there could be large differences between biological samples (e.g. different humans) or should they be much more similar than between treatments (e.g. different types of wine)?
Cordial greeting. I already run two groups of samples (control and treatment, four biological samples in each group ) in XCMS online. It gives me a table of results, telling me what are the metabolites differencially expresed between both set of groups. However if I look for example the m/z given by XCMS I cant find thit in my real data, so,
Is XCMS doing a mean between the four biological samples within groups?
How can I deal with this, I definitively need to compare both groups.
Thanks