A SMILES-based molecular classification platform for screening compounds with potential activity against Dengue, Zika, and Chikungunya — using RDKit descriptors and a Machine Learning model trained for this classification task.
$ cd backend $ python3 -m venv venv $ source venv/bin/activate $ pip install -r requirements.txt $ cd .. $ bash start.sh ✓ Backend http://localhost:8000 ✓ Frontend http://localhost:3000
C:\ArboMol, then open two PowerShell windows in the project folder. We use npm.cmd and venv\Scripts\python.exe so PowerShell's script-blocking policy doesn't get in the way.# Window 1 — backend > cd backend > python -m venv venv > venv\Scripts\python.exe -m pip install -r requirements.txt > venv\Scripts\python.exe main_working.py # Window 2 — frontend > cd frontend > npm.cmd install > npm.cmd start ✓ Opens http://localhost:3000
Everything you need for fast, reproducible molecular screening.
Enter a molecule's SMILES and get the predicted class (Dengue, Zika, or Chikungunya) in seconds.
Upload a CSV or Excel file with multiple molecules and download the processed results as a spreadsheet.
Every SMILES is validated via RDKit, including a molecular weight check (200–700 Da) before prediction.
Hundreds of molecular descriptors automatically computed to feed the classification model.
FastAPI backend with interactive documentation (/docs), ready to integrate with other tools.
Spin up backend and frontend with a single command (docker compose up), no manual configuration.
Four simple steps to run the platform on your computer.
Download the file using the button above and extract the folder anywhere on your computer.
Create the Python virtual environment and install the dependencies: python3 -m venv venv && pip install -r requirements.txt
From the project's root folder, run bash start.sh — it starts the backend and frontend together.
Open http://localhost:3000 in your browser and start classifying, individually or in batch.
ArboMol Classificator sits at the intersection of cheminformatics and machine learning applied to natural product and bioactive compound research. The goal is to make the initial screening of molecular candidates faster and more accessible, indicating a molecule's potential activity against three arboviral diseases of major public health relevance in Brazil: Dengue, Zika, and Chikungunya.
The platform computes molecular descriptors from SMILES notation using RDKit and applies a trained Machine Learning model to indicate the most likely class, serving both as a research support tool and as an educational resource for chemistry students and related fields.
Like any computational screening tool, the results are indicative and do not replace experimental validation — the goal is to prioritize candidates for further investigation, not to issue diagnoses.
Download ArboMol Classificator and run the platform on your computer in just a few minutes.