Minds.ai Netrin - AI-based molecular property pred

Minds.ai Netrin - AI-based molecular property pred

"Minds.ai’s Molecular Property Predictor and Molecule Generator-Netrin is an AI-based tool for exposing the structure:property relationships of molecules. It uses graph convolutional deep neural networks to make predictions of molecules’ properties such as toxicity, solubility, lipophilicity, reactivity with other compounds, protein binding affinities, etc.. Input your desired chemical properties into Netrin’s Molecule Generator and receive a dataset of novel molecules and predictions for how well these molecules will match your specified properties. From pharmaceuticals to chemical manufacturing, Netrin accelerates the molecular screening and synthesis process by predicting the outcomes of wet lab tests in seconds compared to years. It also allows a parallel approach for multi-variate optimizations, which is not feasible with traditional workflows. Netrin is a toolchain built on top of TensorFlow and RDKit. It supports standard, spectral and spatial graph convolutional layers.

*Please note that member solutions are often customizable to meet the needs of individual enterprise end users.



  • Molecular Property Prediction
    Molecular property prediction and molecule generator for pharmaceutical and chemical screening and synthesis
  • Scalability
    The system can easily distribute training workloads over multiple computational resources. Intel® Math Kernel Library (MKL) and Intel® Xeon Scalable processors help drive their incredibly fast neural network training
  • Speedup Drug Discovery & Realize TCO
    Reduce research and design time and cost for new pharmaceutical and industrial chemicals .


WorldwideIntel® Xeon ScalableTensorFlowHealthcareManufacturingIntel® Distribution for PythonIntel® Math Kernel Library (Intel® MKL)Intel® Math Kernel Library for Deep Neural Networks (Intel MKL-DNN)ProprietaryDrug DiscoveryDeep Learning