Overview

ModelDB is a computational neuroscience resource seeking to facilitate discovery of modeling work, analyses of the work, and meta-analyses of the relationship between models.

This is the development repository for the redesign of ModelDB intended to be launched in summer 2020.

Primary goals of this redesign include:

• Facilitating analysis: with built-in tools and easier data querying.
• Improved community engagement: using standard technologies and ontologies to make it easier for others to contribute and for us to interoperate with the broader neuroinformatics community.
• Support on more devices: the resource should be useful to people on mobile devices, desktops, and anything in between.

Pull requests are welcome. Please make an issue first so we understand the problem solved or feature being added by the pull request.

Bootstrapping system and data

This repository is currently intended to be bootstrapped from the classic EAV/CR version of ModelDB. One approach to do so follows:

• Download the bitnami django stack OVA

• Import it into a virtual machine such as VirtualBox

• Install the community edition of MongoDB following these directions

  • note: if you're using Amazon lightsail instead, you'll need to switch to the instructions for Ubuntu. At this writing it's using Ubuntu 16.04.

• Create a MongoDB account with readwrite access to a specific database (for simplicity, you may want to call it modeldb)

  • launch MongoDB by typing mongo
  • switch to the database you want to use e.g. use modeldb

  • create the user:

    db.createUser({ user: "username", pwd: "password", roles: [{role: "userAdmin", db:"modeldb"}]})

• Install git so you can clone this repository, if it's not already installed sudo apt install git

• Clone this repository

• Install Python dependencies (sudo pip3 install -r requirements.txt)

• Create four directories for storing (1) the model zip files, (2) private model zip files, (3) unprocessed references, (4) modelview data.

• Create a settings file in the exact path /home/bitnami/modeldb-settings.json. It should have values for:

  • secret_key (I'm unclear on if there are any rules on this, but I guess a random 50 character string should work)
  • mongodb_user
  • mongodb_pw
  • db_name -- set this to modeldb or whatever you called the database you wish to use.
  • modeldb_zip_dir
  • modeldb_private_zip_dir
  • unprocessed_refs_dir
  • modelview_dir
  • metadata-predictor-rules (points to a file with predictor rules; you can use the one in resources)
  • smtp_server (optional, but required for sending mail)
  • smtp_user (optional, but required for sending mail)
  • smtp_password (optional, but required for sending mail)
  • news (optional, but a file listed here will be used for filling out the news section of the homepage)

• Run the scripts in the extract_data directory in this order:

  • init_db.py
  • extract_data.py
  • note: there are a large number of data items (especially papers). This will take a while.
  • get-zips.py
  • setup_cell_links.py
  • setup_trees.py
  • get_icg_mapping.py
  • get_file_hashes.py
  • this generates the file reuse detection data
  • sync_neuron_dict.py
  • this updates the neurons included in celltypes

• Apply the django migrations python3 manage.py migrate

• You will also want to use django admin to create a user with admin permissions from within the shell you get via python3 manage.py shell. Example from: https://docs.djangoproject.com/en/3.0/topics/auth/default/#creating-users

  from django.contrib.auth.models import User
  user = User.objects.create_user('john', 'lennon@thebeatles.com', 'johnpassword')

• Pull links to biosimulations: python3 pull_biosimulations.py (in the extract_data folder). The website will still work without this, you just won't get any links to biosimulations.org

• Insert static links to Open Source Brain: python3 pull_opensourcebrain.py (in the extract_data folder)

• You can run a development server via, e.g. python3 manage.py runserver 8888

  • This will make the website available on port 8888 (you can then access it from your host system via port-forwarding.
  • This is separate from apache, which is also running and can later be connected to your django system.

On the pipeline

• This requires some files copied over from https://github.com/mcdougallab/pipeline
  • cp pipeline/Project/pipeline_* ~/apps/django/django_projects/Project/Project/
  • cp -r pipeline/Project/templates/pipeline/ ~/apps/django/django_projects/Project/Project/templates/
• Also, note the authentication will need to be setup to use the different database

Deployment hints

• If you're deploying on bitnami's django stack, see their instructions at: https://docs.bitnami.com/virtual-machine/infrastructure/django/get-started/deploy-django-project/
• be sure to turn off debugging in the settings file
• the sqlite3 database needs to be writeable and it needs to be in a folder that's writeable (so not in a path that hosts the website code) (e.g. you might put it in /home/bitnami/db/ and modify settings.py accordingly)
• make sure the folders with private-zips and unprocessed refs can be written to by the server
• if you make any changes on a bitnami machine with apache setup; run sudo /opt/bitnami/ctlscript.sh restart apache to restart
• wsgi.py needs the correct name of the settings module... e.g. if you put it in Project instead of modeldb2020, you should change modeldb2020.settings to Project.settings
  • the settings file may need the full path to TEMPLATES["DIRS"]

on Analysis

The analysis folder contains scripts for analyzing ModelDB in ways that may or may not be supported in the web interface. It includes:

• metadata_repeat_list.py -- provides a list of all used metadata; combine with a free word-cloud generator to visualize

On ModelViews

You can gather a morphology suitable for frontpage display via

result = []

for sec in h.allsec():
    xs = [sec.x3d(i) for i in range(sec.n3d())]
    ys = [sec.y3d(i) for i in range(sec.n3d())]
    zs = [sec.z3d(i) for i in range(sec.n3d())]
    ds = [sec.diam3d(i) for i in range(sec.n3d())]
    arcs = [sec.arc3d(i) for i in range(sec.n3d())]
    result.append([xs, ys, zs, ds, arcs])

but note that in order to color by segment, you need each piece of result to correspond to a segment instead of to a section.

Technologies

The ModelDB website is powered in part by a number of other technologies, including:

Backend

• python
• django
• mongodb
• pymongo
• bitnami django stack

Frontend

• jQuery
• jQuery UI
• Bootstrap
• Font Awesome
• plotly
• select2
• popper
• Bootstrap Table

Did we miss something from this list? Submit a pull request!