I have an issue where converting Thermo Vanquish LC- Q/E Orbitrap MS raw files into mzML files and compressing with zlib compression still isn’t enough to make the files smaller size to upload for spectral processing. So, I used continuous wavelet transform in MSConvert to peak pick in order to decrease data size. That did work and now my data is small enough to upload. However, won’t this cause issues if I try to spectral process since the spectral processing already performs peak-fitting? My collaborator has already tried to use R to do the processing, but it keeps stopping because the data is too big. My peak-picking fix is the only thing that seems to be working with this data so far. What would you recommend? Is it an issue if I upload peak-picked data to spectral processing or will it just pick the same peaks? I want to do statistical analysis, but I have not been able to extract the features yet. And XC-MS is down so I can’t use that either. Thoughts?
Hi,
The current spectral sample size limit is 200MB per spectra after zipping. For most cases, if you centroid them correctly, they will be much less than 100MB. Please double check. If you are still not able to compress the file into the size limit because it is truly too big, you can try to use our Pro-MetaboAnalyst, which offers more data uploading approach. Have a look at this link.
Or you can share the dataset with me and I can have a look when I get a chance.
Bests,
Zhiqiang
Thanks for the response! I was able to centroid them with continuous wavelet transform and they are much smaller now (around 30-40 MB). I am really new at this and was curious if doing the peak picking by CWT before spectral processing is appropriate or not since the spectral processing algorithm also does peak-picking (if I am understanding it correctly). Is it okay to do CWT before spectral processing in MetaboAnalyst? I just wasn’t sure if it did it twice if that is a problem. Thanks!