Skip to content

Latest commit

 

History

History
59 lines (32 loc) · 2.32 KB

README.md

File metadata and controls

59 lines (32 loc) · 2.32 KB
Raw

singnetlogo

Gene Annotation service

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

  • Current annotation sources include Gene Ontology, Reactome pathway database, and BioGRID protein-protein interaction database.

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

"gene annotation UI screenshot"

Local Setup & Development

  1. Download the datasets from https://mozi.ai/datasets

    Note: The datasets are compressed and you have to extract the files before using them.

  2. Create datasets folder in your home directory and move the extracted datasets to that directory

  3. Create results folder in your home directory and define a RESULT_DIR environment variable that points to the directory. This is where the scheme results will be stored.

    cd $HOME
mkdir results 
export RESULT_DIR=$HOME/results

  4. Define the following Environment variables.

     $ export SERVICE_ADDR=localhost

 $ export GRPC_ADDR=http://localhost:3001

 $ export RESULT_ADDR=http://localhost:3002
 
 $ export DATASET=$HOME/datasets

  5. Clone the Project and checkout to the development branch.

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

  6. Build & Run the atomspace-rpc server. You can get the instructions on the githup repo here.

    • Make sure you use --host 0.0.0.0 option while starting the server
    • If you use a custom port (other than 50051), make sure to update the port in the ATOM_SERVER env variable in docker-compose-dev.yml

  7. Go the project directory and Start the services

     docker-compose -f docker-compose-dev.yml up --build

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

    http://localhost:3003