FYI errors in peak picking can be solved
Gavin Blackburn
Hi all,
This is an FYI. I’ve been having some problems with some data sets recently. During the XCMS peak picking stage I’ve been getting error messages that look like this:
Error in .local(object, ...) : m/z sort assumption violated !
Usually with some values included to tell you where the error is.
If you search online you’ll fine mention of trying to remove zero values. This does not seem to be the problem and won’t work.
They are caused by extremely small peaks in the data set bugging the converter. There is a fix, it involves adding a threshold level to the script that is used to convert your files.
If you are happy editing the scripts the filter that you need to add is this:
--filter \"threshold absolute 10 most-intense\"
This will remove anything that is less than 10 counts in your data (value chosen for Thermo instruments).
If you are not happy editing them and are seeing this error please get in touch and I’ll help you.
Cheers,
Gavin.
Dr. Gavin Blackburn
SULSA Technologist
Strathclyde institute of Pharmacy and Biomedical Science
161 Cathedral Street,
Glasgow.
G4 0RE
Tel: +44 (0)1415483828
Metabolomics Society Meeting 2013 – Glasgow. Please download the flyer below:
ScotMet: The Scottish Metabolomics Facility