Webinar: AIMNet2 – A Robust Neural Network Potential for Complex Organic and Element-Organic Chemistry
In this second webinar of the AMS 2024 webinar series, Dr. Roman Zubatyuk from the Lab of Prof. Isayev introduces AIMNet2, an advanced neural network potential designed to transform the modeling of complex organic and element-organic molecules.
The AIMNet2 model is engineered for efficient and accurate simulations, combining machine learning parameterization for short-range interactions with physics-based long-range electrostatic and dispersion terms. It addresses the limitations of traditional quantum mechanical methods by offering a scalable, high-throughput alternative with minimal loss in accuracy.
Roman will discuss AIMNet2’s unique features, including its ability to handle:
- Neutral, charged, organic, and element-organic molecules: learn how the AIMNet2 model excels in describing various molecules, from simple organic compounds to complex element-organic structures.
- Accurate prediction of chemical properties: explore how AIMNet2 ‘s multi-task learning capabilities enable precise predictions of various chemical properties, including energy, forces, and partial charges.
- Efficient exploration of potential energy surfaces: understand how AIMNet2 integration with the Amsterdam Modeling Suite (AMS) facilitates efficient exploration of potential energy surfaces, enabling insights into complex molecular behavior.
- Applications in conformational search, geometry optimization, and more: discover how AIMNet2 has been successfully applied in various tasks, including conformational searches, geometry optimization of small to large molecules, and even molecular dynamics simulations.
Whether you’re a seasoned computational chemist or new to the field of machine learning potentials, this webinar will provide valuable insights into AIMNet2’s capabilities and its potential to accelerate your research.
Webinar Details:
Date: Tuesday 19.11.2024
Time: 17.00 (CET) / 11.00 (ET)
Save Your Spot:
