MAGI is a tool to quickly find and score consensus between metabolite identifications and gene annotations. Upload FASTA and metabolite files here to begin.
If you found MAGI to be useful for your research, please use the following citation:
Onur Erbilgin, Oliver Rübel, Katherine B Louie, Matthew Trinh, Markus DeRaad, Tony Wildish, Daniel Udwary, Cindi Hoover, Sam Deutsch, Trent R Northen, Benjamin P Bowen. MAGI: A Method for Metabolite Annotation and Gene Integration., ACS Chem Biol. 2019 Apr 19;14(4):704-714. doi: 10.1021/acschembio.8b01107. DOI: 10.1021/acschembio.8b01107 Epub 2019 Apr 4. PMID: 30896917.
Your Job ID is a2b72f05-1cb5-4e3b-b77b-6e71f5dadf5d
Your Job URL is magi.nersc.gov/jobs/?id=a2b72f05-1cb5-4e3b-b77b-6e71f5dadf5d
Running MAGI jobs at NERSC is no longer being supported. Please either download the code and run it yourself or run the job in GNPS.
MAGI jobs (MAGI2) can be run at GNPS at the following URL: https://proteomics2.ucsd.edu/ProteoSAFe/index.jsp
The Job submission for MAGI2 is slightly different than what is described here for MAGI1.
Go to the GNPS url above and Select the workflow "MAGI2", upload a protein fasta file, upload a tab delimited text file for the compound file using SMILES as the "original_compound" attribute and "compound_score" as a weight attribute. Any other compound metadata you desire can be uploaded as well.
The benefit of MAGI2 is that the SMILES pattern reduces the sensitivity/specificity to which compound you are asserting.
FOR MAGI1, with the public release of MAGI, many users have noticed extreme sensitivity of MAGI to the exact inchi-key used for original_compound. The table of inchi-keys that we use can be found here https://magifiles.nersc.gov/processed/unique_compounds.csv.gz.
We hope to have a generalized solution to these problems soon and apologize for any inconvenience this may cause.