Untargeted pipeline demo (WIP)

[joewandy]

PR to add demo using vimms to generate synthetic data so we can test the performance of a typical untargeted data processing pipeline in LC/MS metabolomics.

So we generate chemicals in various experimental setup, write their mzml files + corresponding ground truth. The mzml files are then passed into a pipeline that tries to infer back chemical identities from spectral data. Since it's synthetic data, we can easily evaluate pipeline performance for all these steps:

1. Peak picking
2. Alignment
3. Group related peaks like isotopes and adducts together.
4. Identification, by matching to standards / spectral library.

Still work in progress, do not merge yet!!