CarcaraQSAR

CarcaraQSAR - Computational Algorithms to Relate Chemical Attributes with Respective Activity

CarcaraQSAR is an open-source, web-based framework designed to simplify the development of Quantitative Structure-Activity Relationship (QSAR) models. By integrating machine learning algorithms and bioinspired feature selection techniques, the platform enables researchers to efficiently identify chemical descriptors correlated with biological activity.

Its user-friendly interface eliminates the need for extensive programming knowledge, making QSAR modeling more accessible. CarcaraQSAR supports multiple validation strategies, including cross-validation and Y-randomization, ensuring robust and reproducible models. The platform is scalable, allowing cloud deployment, and is particularly suited for applications in drug discovery and computational chemistry.

Demonstration Video

Click the link below to watch the demonstration video.

https://www.youtube.com/watch?v=ALBr-ow9J_o

What is a QSAR Model?

A QSAR (Quantitative Structure-Activity Relationship) model is a computational technique used in chemistry and pharmaceutical sciences to predict the biological, chemical, or toxicological properties of compounds based on their molecular structure. These models are essential in fields like drug design, toxicology, and environmental chemistry.

The general steps for building a QSAR model include:

Data Collection: Gathering molecular structure and biological activity data.
Feature Selection: Identifying relevant molecular descriptors such as molecular weight, charge distribution, and functional groups.
Model Development: Using statistical or machine learning techniques (e.g., regression, decision trees, neural networks) to create predictive models.
Model Validation: Evaluating performance using validation techniques like cross-validation and Y-randomization.
Application: Using the model to predict properties of new chemical compounds.

CarcaraQSAR automates these steps, making QSAR modeling more accessible and efficient.

Installation Guide

CarcaraQSAR can be installed and executed locally on both Linux and Windows.

Linux Installation

Prerequisites

Ensure the following dependencies are installed before proceeding:

Git
Python + pip
Node.js + npm
Redis

To install Redis, run:

sudo apt install redis-tools redis-server

Installation Steps

Clone the repository:

git clone https://github.com/rosalvoneto/CarcaraQSAR
cd CarcaraQSAR

Frontend Setup
```
cd frontend
npm install
npm run dev
```

Backend Setup (Open a new terminal in the current path)

cd backend/backend
python3 -m venv venv
source venv/bin/activate  # Activate virtual environment
pip install -r requirements.txt
python3 manage.py migrate
python3 manage.py runserver 0.0.0.0:8000

Start Redis and Celery (Open a new terminal in the current path)
```
redis-server &
celery -A backend worker -l info
```
Open your browser and access:
```
http://localhost:5173/
```
Create an account and start using CarcaraQSAR.

Windows Installation

1. Instalar o WSL

Se você ainda não tem o WSL instalado, siga esses passos:

Habilite o WSL e o Virtual Machine Platform:
- Abra o PowerShell como Administrador e execute o seguinte comando:
```
 wsl --install
```
Escolha a distribuição Linux:
- O WSL instalará o Ubuntu por padrão. Para confirmar se o Ubuntu foi instalado, execute:
```
 wsl --list --verbose
```
Reinicie o PC para completar a instalação.
Após o reboot, abra o Ubuntu (ou a distribuição escolhida) e siga as instruções para configurar o nome de usuário e senha.

2. Criação do script `setup_project.sh`

Copie o script abaixo para a sua área de transferência.

#!/bin/bash

echo "======================="
echo "Setup do CarcaraQSAR"
echo "======================="

# Atualizar e instalar dependências do sistema
echo "Atualizando e instalando pacotes do sistema..."
sudo apt update
sudo apt install -y python3 python3-venv python3-pip redis-server nodejs npm git libpq-dev

# Iniciar Redis se ainda não estiver rodando
if pgrep -x "redis-server" > /dev/null; then
    echo "Redis já está rodando."
else
    echo "Iniciando Redis..."
    sudo service redis-server start
fi

# Clonar o repositório apenas se ainda não existir
if [ ! -d "CarcaraQSAR" ]; then
    echo "Clonando o repositório CarcaraQSAR..."
    git clone https://github.com/rosalvoneto/CarcaraQSAR
else
    echo "Repositório CarcaraQSAR já existe. Pulando clone."
fi

cd CarcaraQSAR

# Configuração do frontend
echo "Configurando o frontend..."
cd frontend
npm install

if pgrep -f "vite" > /dev/null; then
    echo "O servidor frontend (Vite) já está rodando."
else
    echo "Iniciando o frontend (Vite)..."
    npm run dev &>/dev/null &
fi

# Voltar para o diretório raiz do backend
cd ../backend/backend

# Criar e ativar o ambiente virtual se necessário
if [ ! -d "venv" ]; then
    echo "Criando ambiente virtual..."
    python3 -m venv venv
fi

echo "Ativando o ambiente virtual..."
source venv/bin/activate

# Instalar as dependências do Python
echo "Instalando dependências Python..."
pip install -r requirements.txt

# Rodar migrações do Django
echo "Aplicando migrações do Django..."
python3 manage.py migrate

# Rodar o servidor Django se não estiver rodando
if pgrep -f "runserver" > /dev/null; then
    echo "O backend Django já está rodando."
else
    echo "Iniciando o backend (Django)..."
    python3 manage.py runserver 0.0.0.0:8000 &>/dev/null &
fi

# Rodar o Celery se não estiver rodando
if pgrep -f "celery" > /dev/null; then
    echo "O Celery já está rodando."
else
    echo "Iniciando o Celery..."
    celery -A backend worker -l info &>/dev/null &
fi

# Conclusão
echo "Configuração concluída!"
echo "Frontend: http://localhost:5173/"
echo "Backend (Django): http://localhost:8000/"

3. Passo a Passo para Usar o Script

Crie o arquivo setup_project.sh no diretório onde deseja rodar o projeto: No terminal do WSL, no diretório onde você quer que o projeto seja instalado, execute:
```
 touch setup_project.sh
 nano setup_project.sh
```
Cole o conteúdo do script acima no arquivo setup_project.sh.
Torne o script executável:

No terminal do WSL, execute:
```
 chmod +x setup_project.sh
```
Execute o script:

Ainda no terminal do WSL, execute o script com o seguinte comando:
```
 ./setup_project.sh
```

4. Entrar no CarcaraQSAR

Agora, basta abrir o link http://localhost:5173/ no navegador e explorar o projeto!

Demonstration

For a quick demonstration, an online version of CarcaraQSAR is available. However, note that this instance has limited hardware resources and should not be used for large-scale projects.

Access the Demo

To request access, please fill out the form below:

👉 Access the Demo Form

To test the platform, use the dataset F25_S50_YL.csv, available in the root directory of the GitHub repository.

License & Citation

CarcaraQSAR is licensed under CC BY-NC (Creative Commons Attribution-NonCommercial). If you use this software in your research, please cite:

CarcaraQSAR: A Computational Framework for Relating Chemical Descriptors to Biological Activity
Daniel Alencar Penha Carvalho, Rosalvo Ferreira de Oliveira Neto, Edilson B. Alencar Filho
To be submitted to SoftwareX.

For more details, visit the GitHub repository.

This site is open source. Improve this page.