Replace homemade annotation script with W4M's `lcmsmatching`

[elliotfontaine]

https://github.com/workflow4metabolomics/lcmsmatching

Steps:

• [ ] Test lcmsmatching with our data and discover its interface.
• [ ] Compare results.
• [ ] Compare speed.
• [ ] Implement it in a task, with additional table formatting if necessary.

[elliotfontaine]

See discussion in https://github.com/workflow4metabolomics/lcmsmatching/issues/68.

Replacing the mzmatch junk script used by our workflow by the lcmsmatching R script won't be straightforward. The latter doesn't work with current versions of its core dependency biodb. One approach would be to partially reimplement lcmsmatching using biodb 1.10.0, and following what's inside vignettes:

• https://www.bioconductor.org/packages/release/bioc/html/biodb.html
• Mass spectra annotation with a compound database
• Accessing a custom CSV file with a biodb connector