Biocentral server

Flask-based python server using REST API that provides functionality for the biocentral frontend.

Features and modules

These are the currently provided modules:

Module	Features	Dependencies
protein_analysis	wip
embeddings	Embed protein sequences	biotrainer
ppi	Work with protein-protein interactions	hvi_toolkit
prediction_models	Train and evaluate models trained on embeddings	biotrainer
proteins	Work with protein data	taxoniq
plm_eval	Automatically evaluate protein language models	autoeval, FLIP

Supported platforms

Biocentral and biocentral_server are designed to provide all common desktop operating systems (Windows, Linux, macOS). The following table gives an overview about the current test and packaging status:

OS	Tested	Packaging
Ubuntu 24.04	✅	.zip
Ubuntu 22.04	✅	.zip
Windows 10	✅	.zip
Windows 11	planned	tbd
macOS	planned	tbd

Installing and running

Make sure that you have Python 3.11 and poetry installed.

# [Ubuntu 24.04] 
# Install additional dependencies 
sudo apt-get install python3-tk
sudo apt-get install libcairo2-dev libxt-dev libgirepository1.0-dev
poetry install --extras linux

# [Windows 10/11]
poetry install
# Install torch with hardware settings for your system (see here: https://pytorch.org/get-started/locally/)
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

# Run with visual control panel
poetry run run-biocentral_server.py

# Run headless
poetry run run-biocentral_server.py --headless

Additional setup

Embeddings that are computed via the server are stored in a NoSQL database. For local deployments, TinyDB is used by default and should be performant for most use cases. If you want to use TinyDB, you do not have to perform any additional installation steps - all embeddings are stored in storage/embeddings.json.

Advanced PostgreSQL setup

For advanced users or production deployments we recommend using a PostgreSQL instance. Here's a step-by-step guide how to configure it for biocentral_server:

# 1. Install PostgreSQL, e.g. for Ubuntu see: https://www.postgresql.org/download/linux/ubuntu/
# 2. Configure PostgreSQL
# Switch to postgres user
sudo -i -u postgres

# Create a new database
createdb embeddings_db

# Access PostgreSQL prompt
psql

# Create a new user and set password
CREATE USER embeddingsuser WITH PASSWORD 'embeddingspwd';

# Grant privileges to the user on the database
GRANT ALL PRIVILEGES ON DATABASE embeddings_db TO embeddingsuser;

# Connect to the embeddings database
\c embeddings_db

# Grant schema privileges to the user
GRANT ALL ON SCHEMA public TO embeddingsuser;

# Exit PostgreSQL prompt
\q

# Exit postgres user shell
exit

# Restart PostgreSQL
sudo systemctl restart postgresql

Building

Building and bundling is done using pyinstaller and make.

On Windows, you can use winget to install make:

winget install ezwinports.make

Check that all build variables are correct:

make print-info

Then call build:

make build

To create a zip file with all required files for distribution:

make bundle

Note that this file only works on the operating system version you ran make build on!

Citation

Please cite the biocentral main repository if you are using biocentral_server in your scientific publication:

@Online{biocentral,
  accessed = {2024-09-10},
  author   = {Biocentral contributors},
  title    = {Biocentral - An open source bioinformatics application},
  url      = {https://github.com/biocentral/biocentral},
}