GT-AutoML

GT-AutoML™ is an automated, tailor-made Structure-Activity Relationship (SAR) modeling platform developed by Standigm, designed to streamline the entire QSAR (Quantitative Structure-Activity Relationship) modeling process for drug discovery teams. Powered by deep learning and chemical intelligence, it supports both ligand-based and structure-based approaches, providing versatility across a wide range of drug discovery programs.

The platform is built to eliminate labor-intensive steps in the QSAR workflow, integrating multi-level molecular descriptors — including inputs from state-of-the-art foundation models — to construct a comprehensive and reliable feature set for predictive modeling.

Core Capabilities

• Automated QSAR Modeling: GT-AutoML™ automates the full QSAR modeling pipeline, reducing manual effort and accelerating model generation for medicinal chemistry and computational biology teams.
• Ligand-Based and Structure-Based Approaches: The platform supports both ligand-based and docking-based QSAR methodologies, enabling flexibility depending on the availability of structural data and the nature of the drug discovery program.
• Automated Ligand Alignment: Labor-intensive ligand alignment steps are handled automatically within the platform, improving throughput and consistency.
• Multi-Level Molecular Descriptors: GT-AutoML™ integrates a broad set of molecular descriptors, including those derived from cutting-edge foundation models, to build a rich and comprehensive feature representation for each compound.
• Deep Learning-Powered Predictions: The platform leverages deep learning to deliver reliable and accurate SAR models, supporting informed decision-making in lead optimization and compound prioritization.

Workflow

1. Input compounds are processed through the GT-AutoML™ decision-tree workflow for SAR model generation.
2. Ligand alignment and descriptor calculation are performed automatically, incorporating multi-level molecular features.
3. Both ligand-based and docking-based QSAR models are constructed and evaluated within the platform.
4. The resulting models provide actionable SAR insights to guide drug discovery programs.

GT-AutoML™ is offered by Standigm, a company focused on AI-driven new drug research and development. A whitepaper is available for teams seeking a deeper technical understanding of the platform's methodology and performance.