Question: Can XCMS output continuum/profile data to netCDF format?
0
2.6 years ago by
ray.bacala0 wrote:

Hello,

I am looking to convert data from my LC-MS (Waters G2 Synapt) to netCDF in order to import into another platform.  Unfortunately netCDF is the only import option available.  Waters Masslynx contains a netCDF export tool, however it is hard wired to centroid the data first and I need non-centroided data (profile/continuum).

I can convert to a suitable import format (mzXML for example) using ProteoWizard but then need to convert to netCDF.  One of the PWiz devs suggested that XCMS can do it but I cannot see this explicitly stated in the package literature.

MAIN QUESTION - If I supply non-centroided data to XCMS, can it translate it to netCDF without performing any processing (centroiding, peak picking).

BONUS QUESTION - I will float this to the ethers, but does anyone know of a better tool to get from Waters.raw to netCDF?

BONUS QUESTION 2 - Is there a Bioconductor app in development that could take the output from XCMS (peak picking), identify ESI-MS protein charge state profiles and deconvolute the mass?  The problem I am having is the identification of protein charge state envelopes from a complex LC-ESI-MS run containing dozens or more protein species without relying on chromatographic integration (many proteins are invisible in a BPI trace, too many co-elutions in a TIC chromatogram and there are abundant proteins under flat baseline in TIC chromatograms.

The main question is what I really need answered, the other two are bonus.

Ray

convert xcms protein • 825 views
modified 2.6 years ago by Johannes Rainer1.4k • written 2.6 years ago by ray.bacala0
Answer: Can XCMS output continuum/profile data to netCDF format?
1
2.6 years ago by
Johannes Rainer1.4k
Italy
Johannes Rainer1.4k wrote:

Indeed, you can load the mzML files as a xcmsRaw object and save that using the write.cdf(object, filename) as a cdf file. Assuming that the mzML file is bound to the mz_file variable:

library(xcms)
library(ncdf4)

## Setting profstep to 0 as we don't need to generate the "profile matrix"
xs <- xcmsRaw(mz_file, profstep = 0)
write.cdf(xs, filename = sub(mz_file, pattern = ".mzML", replacement = ".cdf", fixed = TRUE))

cheers, jo

Hello,

Thank you for your prompt answer.  From the looks of the above script, I assume that if the target software could not read netCDF4 for some reason, I would be able to obtain a prior version library for netCDF and call that in?

Ray

I don't quite understand your question. What target software do you refer to? xcms uses the R package ncdf4 to write the data. I would assume that most software should be backward compatible and be able to read netCDF4.

cheers, jo

Hello,

I am trying to go from Waters ,raw to import to Bruker Compass.  What I did not know is what VERSION of netCDF Compass could read.  I initially was concerned I might have to write to netCDF 2 or something BUTI have since found out from Bruker that they can read netCDF4, so I will be fine.

The reason I have to do all of this is because while Waters has a netCDF converter (DBridge), it seems to automaticallty centroid ToF data during the conversion.  I am talking with Waters about that now, but if they cannot change it, I will have to use ProteoWizard to convert .raw to .mzML and then XCMS to convert mzML to netCDF.  Now if XCMS could call in the Waters data library directly (they have one I can access and I think it is the same one that ProteoWizard is using), I could do the conversion in one step...

Ray

no, xcms can not read the raw data files. as far as I know proteowizard can do that for some manufacturers on certain platforms. Building that support into xcms is not possible.

Hello,

I finally have had all of the software I need installed and tried the conversion out.  To start, XCMS choked on my mzML conversion and on reading I saw no mention of this format being supported.  No big deal, I converted to mzXML instead using ProteoWizard.

I was able to then convert to netCDF with XCMS but the Bruker software would not open it, stating it was not netCDF.  After two hours of assuming I was the problem (the probable bet), I located this text on page 69 of the XCMS manual:

"Currently the only application known to read the resulting file is XCMS. Others, especially those which build on the AndiMS library, will refuse to load the output."

Does this mean that XCMS netCDF format is not really netCDF format?  Is there something I am missing?

Regards,

Ray