This script downloads results of spectral annotations from classical molecular networking or feature-based molecular networking job from GNPS [http://gnps.ucsd.edu] and generate virtual metabolites either with SyGMa and/or BioTransformer. The resulting candidates can be used for Network Annotation Propagation on GNPS or with SIRIUS.
Click on the following link to launch the vm-NAP web-app. Note that this is a temporary instance with limited ressources. For better performance or larger projects, please install and use it locally.
Install locally in conda with:
Download the present repository: Use git clone or download manually.
In the terminal, navigate to the repository folder.
Install the environment with:
conda env create --file environment_local.yml
Initiate the environment:
conda activate vm-NAP
For using the same interface as the website, use the Streamlit web-app locally with the following steps:
Be sure the app is installed localy and you are in the repository folder.
Start the streamlit app with:
streamlit run vm_NAP_streamlit.py --server.port 8501 --server.address 0.0.0.0
or
streamlit run vm_NAP_streamlit.py --server.port 8501 --server.address localhost
Representative command for the python script:
python src/vm_NAP_processing.py --job_id=bbee697a63b1400ea585410fafc95723 --run_sygma --run_biotransformer --sirius_input_file=input/compound_identifications.tsv --debug --max_compounds_debug=3
For help in how to use it, run:
python src/vm_NAP_processing.py --help
See the documentation for custom database in NAP and how to run Network Annotation Propagation (NAP) on GNPS https://ccms-ucsd.github.io/GNPSDocumentation/nap/#structure-database.
See the documentation to generate the SIRIUS custom database here.
IMPORTANT: Note that only spectral annotations that have a valid InChI or SMILES identifier will be considered downstream. If the annotations you are interested in don't have an identifier in the library, go back to the GNPS library entry, update the entry by adding an identifier, and rerun your GNPS job