Getting Started

Overview

Biosimulation Model Search Engine (BMSE) is a web-based search engine for finding information in biosimulation models created using CellML and stored in the Physiome Repository Model (PMR). Types of information include variables, mathematical equations, constants, components, models, images, and simulation results. This work uses Composite Annotation Search Using BERT (CASBERT) to represent queries and entities in biosimulation models [MRN22].

Deployed Search Engine

BMSE: http://search.bm-se.cloud.edu.au/

Documentation

Requirements

Local installation using Docker

  • Make sure Docker is installed

  • git clone the project

    git clone https://github.com/napakalas/bmse.git

  • change current directory to bmse

    cd bmse

  • build/start BMSE

    make start

  • Now you can access bmse via web browser with

  • Check the Makefile for commands such as stop services, view container logs, etc.

How to deploy?

We deploy BMSE on a cloud service by NECTAR. Services run on an Ubuntu 18.04 LTS (Bionic) amd64 instance with 4 VCPUs and 8GB RAM. We utilise ‘docker context’ for deployment, which provides flexibility in selecting deployment targets. Here are the steps:

  • install Docker on the target deployment instant (please adjust it for

    other than Ubuntu 18.04 LTS (Bionic)):

    • Connect to the target deployment instant:

      ssh ubuntu@TARGET-PUBLIC-IP
    • Install docker, follow [the installation instructions]

      (https://docs.docker.com/engine/install/ubuntu/) on Ubuntu.

    • Change docker.sock access permission

      sudo chmod 666 /var/run/docker.sock

    • Logout from the target deployment instant:

      logout

  • build/start BMSE on the remote server

    make start-remote ip=TARGET-PUBLIC-IP

  • now you can access BMSE via web browser with your instance public IP

    http://TARGET-PUBLIC-IP/
  • if you want to implement SSL on the remote server, follow this link:

    https://mindsers.blog/post/https-using-nginx-certbot-docker/