search

Metabolomics-MPC / NTAnnotationWorkflow

An R script for the annotation of small-molecule LC-MS data
2 stars 1 forks source link

Script input #3

Closed michaelwitting closed 2 years ago

michaelwitting commented 2 years ago

I think we can structure the input to the script a bit. In our metabolomics experiments we typically have positive and negative mode data from the same LC method. We are currently working with MetaboAnnotation on functions to match between both ionization modes (Link: #56). It would be nice if we can match at the end of the script the MS1 features between positive and negative ionization mode. Therefore I'm suggesting the following input structure:

|Project folder |param file (containing all tolerances, adducts for positive and negative etc...) |-> pos |--> input |---> xxx.csv (MS1 feature table) |---> xxx.mgf (MS1 and MS2 data) |---> MS1_lib_ext (folder containing compound libraries for MS1 matching using only m/z values) |---> MS1_lib_inhouse (folder containing compound librires for MS1 matching using m/z and RT/RI values) |---> MS2_lib_ext (folder containing MS2 libraries for matching) |---> MS2_lib_inhouse (folder containing MS2 libraries for matching, including RT/RI value) |--> output |->neg |--> input |---> xxx.csv (MS1 feature table) |---> xxx.mgf (MS1 and MS2 data) |---> MS1_lib_ext (folder containing compound libraries for MS1 matching using only m/z values) |---> MS1_lib_inhouse (folder containing compound librires for MS1 matching using m/z and RT/RI values) |---> MS2_lib_ext (folder containing MS2 libraries for matching) |---> MS2_lib_inhouse (folder containing MS2 libraries for matching, including RT/RI value) |--> output

The idea would be to then calc the R script from the command line with input the only input the project folder and then it runs.

What do you think @chufz ?

michaelwitting commented 2 years ago

Shall we have in the yaml file only a path to the MS2 libraries and read files from there. I would iterate over them if multiple libraries shall be used. The file type can identified with a regex on the file ending, e.g. .mb would use the MassBank backen and .msp the MSP backend

chufz commented 2 years ago

Yes i had something like this in mind, you can then decide in the settings file which of the libraries to use

michaelwitting commented 2 years ago

Done.