Apply machine-learning to improve mass spectra based library matching.
Based on: https://github.com/iomega/spec2vec_gnps_data_analysis
An environment to run these notebooks can be set up with something like:
conda create --name spec_analysis python=3.7
conda activate spec_analysis
conda install --channel nlesc --channel bioconda --channel conda-forge spec2vec
pip install jupyter
pip install ms2query