sorenwacker / ms-peakonly Goto Github PK

Hi,

Dependencies have to be installed manually, which is not really optimal.

To help, here is a first list (probably more missing):

mamba install pytorch torchvision torchaudio -c pytorch

mamba install -c bioconda pymzml

mamaba install -c conda-forge bintrees 

Best

Hi,
Very nice work!

Since you did a partial re-write of peakonly and get to know it well, I would have the following question:

In your opinion, what would be the easiest way to use peakonly to detect peaks, not in a 2D chromatogram (retention time, intensity, m/z), but in a 1D chromatogram (retention time, intensity) coming from another detector than MS. Peak shapes should be similar.

I would love to hear from you about it. If my question is unclear for you, I am happy to develop and/or share .mzML files with multiple spectra/chromatogram objects coming from various detectors in it!

Best,

"po = PeakOnly(model_dir='/home/swacker/.peak_only')" in PeakOnly-in-JupyterLab.ipy, I wonder how to generate the .peak_only file. Thank you!

Hi!

I just had the following:

line 843, in convolve
    raise ValueError('v cannot be empty')

It is probably due to https://github.com/Arseha/peakonly/blob/29e9af5bfa0e6b816ff15dcdc903b4c81412177c/processing_utils/matching.py#L320.

One way to fix it is detailed here: https://fixexception.com/numpy/v-cannot-be-empty/

Also tagging @Arseha in case!

Best,

Hi,

Tried a few analyses and looks like peaks are well detected but the retention time does not match at all. It looks like it only gives the very first time the m/z appears (not even in a peak). I have no real idea where it might come from...Might be because of the experimental named tensors?

Would be great! :)

https://pymzml.readthedocs.io/en/latest/index_gzip.html#accessing-the-data