• Accepts list of human HGNC gene symbols, finds annotations from example databases and displays the results as a browser-based interactive graph visualization.

• Results can be downloaded as a JSON file viewable in cytoscape, a set of plain text tables, and an opencog atomese scheme file.

1. Gene Ontology Annotation:

   Find list of Gene Ontology (GO) terms that the gene is a member of, it can also traverse to n number of parent GO's.

2. Pathway Annotation:

   Finds the Small Molecule Pathway Database (SMPDB) and Reactome pathways ID's for a given list of genes and the small molecules and Proteins in the selected pathways.

3. Protein Interaction Annotation:

   Find known interacting proteins from the BIOGRID protein interaction database.

1. Gene Ontology:

Three sets of concepts describing biological processes, chemical functions, and cellular locations associated with gene products, organized as directed acyclic graphs.

2. GO annotation:

Assigns human gene symbols to GO and their corresponding GO term (Gene ID (gene symbol) and GO ID)

3. Reactome:

Curated metabolic and signaling pathways

4. SMPDB:

Includes mostly metabolic pathways with proteins and small molecules.

5. BIOGRID:

Contains experimentally verified protein-protein interactions

1. Clone the Project

   git clone --recursive https://github.com/MOZI-AI/annotation-service.git

2. Define the following Environment variables. $SERVER_ADDR and $SERVER_PORT. If running on a local machine set $SERVER_ADDR to http://localhost

    $ export SERVER_ADDR=<ADDR>
   
    $ export SERVER_PORT=<PORT>
   

3. Start the docker

    docker-compose up
   

4. You should be able to access the annotation service UI at

   http://$SERVER_ADDR:$SERVER_PORT
   

If you would like to run in production mode, which will load large datasets into the atompspace, you can change line number 15 in docker-compose.yml to 1 If you don't want to run the snet kovan network, you can comment lines 18-40 in circus.ini file.