Skip to main content ↑ ↓
Topic: Fold Change Calculation in XCMS (Read 4391 times) previous topic - next topic

Fold Change Calculation in XCMS

Hi All,

Recently I found the XCMSonline has some other ways to calculate the fold change, not just mean1/mean2. For example, in the demo COKE vs Pepsi, the 17th record in the diffreport table. The fold change of that record is 7.235695, however, the mean of dataset2 is 0. I suppose there must be another way to calculate such kind of fold change. Can anyone tell me related method?
Thanks

Re: Fold Change Calculation in XCMS

Reply #1 –
Umm.... I guess you could take the median or really just about any other stat. If there was reason to believe that the distribution was not normal then the median would make sense.

If could just be a difference be due to normalisation. (?) There are other ways of calculating the fold change, including bayesian methods (ref below). I just did a quick pubmed search I'm  sure there are more.

Cheers,

Paul

J Comput Biol. 2001;8(6):585-614.
Bayesian estimation of fold-changes in the analysis of gene expression: the PFOLD algorithm.
http://www.ncbi.nlm.nih.gov/pubmed/11747614
~~
H. Paul Benton
Scripps Research Institute
If you have an error with XCMS Online please send me the JOBID and submit an error via the XCMS Online contact page
 

Re: Fold Change Calculation in XCMS

Reply #2 –
Quote from: "hpbenton"
Umm.... I guess you could take the median or really just about any other stat. If there was reason to believe that the distribution was not normal then the median would make sense.

If could just be a difference be due to normalisation. (?) There are other ways of calculating the fold change, including bayesian methods (ref below). I just did a quick pubmed search I'm  sure there are more.

Cheers,

Paul

J Comput Biol. 2001;8(6):585-614.
Bayesian estimation of fold-changes in the analysis of gene expression: the PFOLD algorithm.
http://www.ncbi.nlm.nih.gov/pubmed/11747614

Hi Paul,

The fold change data is from xcmsonline results. So I guess the xcmsonline use other way to calculate fold change, like bayesian methods as you mentioned?