Team
Predicting Activity in Organic Samples — a SMILES-based platform for screening the biological activity of organic molecules, powered by Machine Learning models trained from molecular docking simulations.
Access the platform directly in your browser — no installation required. Draw or paste a molecule's SMILES notation and get an instant activity prediction.
Platform not loading right away? The backend runs on a free server tier — the first request can take 50 seconds to 1 minute while it wakes up.Everything you need for fast, reproducible molecular activity screening.
Draw structures or paste SMILES directly using the built-in Ketcher chemical editor — no external software needed.
Get a clear ATIVO (active) or INATIVO (inactive) classification against the target in seconds.
The underlying ML model was trained on descriptors derived from molecular docking simulations, not experimental data alone.
Inspect the predicted molecular structure in an interactive 3D viewer right in the browser.
Skip hours of docking setup — PREDACTORS delivers a fast, indicative prediction using pre-trained models.
Built for researchers, students, and chemists — free to use for educational and research purposes.
No installation and no docking software — everything runs in your browser.
Open PREDACTORS at predactors.vercel.app — no installation needed.
Draw the molecule or paste its SMILES notation into the input field.
The system loads the pre-trained ML model and computes molecular descriptors — this can take 50 seconds to 1 minute.
The molecule is classified as ATIVO or INATIVO based on its predicted interaction with the biological target.
PREDACTORS was developed as part of research exploring the molecular potential of natural products against COVID-19 through molecular modeling and Machine Learning. The platform makes the resulting predictive model publicly accessible to other researchers, chemists, and students.
By combining molecular descriptors computed via cheminformatics tools with a Machine Learning model trained on molecular docking simulation data, PREDACTORS estimates whether a given organic molecule is likely to be active against a biological target — without requiring users to run their own docking software.
As with any computational screening tool, predictions from PREDACTORS are indicative, not diagnostic. They are meant to help prioritize candidate molecules for further experimental or computational investigation, not to replace laboratory validation.
Access PREDACTORS and get an activity prediction for your compound in less than a minute.
