CoviPredX: Coronavirus Bioactivity Prediction using XGBoost
Key Features
Accurate pIC50 Prediction: CoviPredX predicts the pIC50 values, assisting the researchers in evaluating potential bioactivity against coronavirus using an advanced XgBoost regression algorithm.
Molecular Fingerprint-Based Predictions: The tool leverages molecular fingerprints as input features for robust bioactivity prediction, enhancing drug discovery efforts.
Supports Canonical SMILES Input: CoviPredX accepts Canonical SMILES format for small molecules, making it easy to input chemical structures directly into the app.
User-Friendly Interface: The web interface is simple to use and allows the users to upload the datasets in CSV format for immediate prediction.
Cross-Platform Availability: The app is a web-based tool on Linux and Windows, ensuring cross-access for different environments. Click on the GitHub link for the standalone version of the app.
Why Use CoviPredX?
Fast and Reliable Predictions: Quick response for results in seconds with a highly efficient XgBoost model.
Drug Discovery Support: Ideal for researchers working on COVID-19 therapeutics, helps to narrow down potential drug candidates faster.
Input Requirements
SMILES Format: Standard input for chemical compounds.
CSV File: The app requires a CSV file containing SMILES and IDs for prediction.
Output
pIC50 Values: Predicts bioactivity values for the uploaded molecules and evaluates the potential effectiveness against COVID-19.