Analysis and benchmarking of mass spectra similarity measures using gnps data set.
If you use spec2vec for your research, please cite the following references:
F Huber, L Ridder, S Verhoeven, JH Spaaks, F Diblen, S Rogers, JJJ van der Hooft, "Spec2Vec: Improved mass spectral similarity scoring through learning of structural relationships", bioRxiv, https://doi.org/10.1101/2020.08.11.245928
(and if you use matchms as well: F. Huber, S. Verhoeven, C. Meijer, H. Spreeuw, E. M. Villanueva Castilla, C. Geng, J.J.J. van der Hooft, S. Rogers, A. Belloum, F. Diblen, J.H. Spaaks, (2020). matchms - processing and similarity evaluation of mass spectrometry data. Journal of Open Source Software, 5(52), 2411, https://doi.org/10.21105/joss.02411 )
Thanks!
Possibly the easiest way to learn how to run Spec2Vec is to follow our tutorial on matchms and Spec2Vec.
- Part I - Import and process MS/MS data using matchms
- Part II - Compute spectral similarity using Spec2Vec
- Part III - Create molecular networks from Spec2Vec similarities
Current spec2vec works with Python 3.7 or 3.8, it might also work with earlier versions but we haven't tested.
conda create --name spec2vec_analysis python=3.7 # or 3.8 if you prefer
conda activate spec2vec_analysis
conda install --channel nlesc --channel bioconda --channel conda-forge spec2vec
pip install jupyter
git clone https://github.com/iomega/spec2vec_gnps_data_analysis
cd spec2vec_gnps_data_analysis
jupyter notebook
- Original data was obtained from GNPS: https://gnps-external.ucsd.edu/gnpslibrary/ALL_GNPS.json
- Cleaned and processed GNPS dataset for positive mode spectra (raw data accessed on 2020-05-11), can be found on zenodo: https://zenodo.org/record/3978072
Pretrained Word2Vec models to be used with Spec2Vec can be found on zenodo.
- Model trained on UniqueInchikey subset (12,797 spectra): https://zenodo.org/record/3978054
- Model trained on AllPositive set of all positive ionization mode spectra (after filtering): https://zenodo.org/record/4173596