Genetic Algorithm Driven Molecular Structure Determination and Visualization for Real-Time Decision Support

Bottlenecks for instrument scientists currently lie in computational and data workflow complexities, and a lack of tools that allow real-time refinement to occur while samples are still in the instrument. Using the Genetic Algorithm Driven Decision Support System (GADDSS) the instrument user can steer the analysis as well as the instrument in support of the fastest time to solution. GADDSS aims to utilize parallel HPC resources to aide in the molecular structure determination, refinement, and visualization for the TOPAZ and SNAP instruments at the Spallation Neutron Source by providing a conduit for real-time on-the-fly analysis and experiment decision support. The Orbiter platform will provide federated user authentication and facilitate live data streaming, processing, visualization and storage enabling real-time experiment decision support. Orbiter is comprised of modular, scalable, and secure services that provide a solid foundation for building sophisticated capabilities and user interfaces that can be used to facilitate live data processing and visualization for a samples molecular structure of while it is still in the beam, allowing for enhanced experiment control and beam time optimization. We propose that this platform can be extended to other instruments and DOE BES facilities to provide real- time decision support for facility users, and can be made available commercially as a robust experiment data analysis and visualization application. The primary success of the GADDSS project was the implementation of a robust experiment data process- in, analysis, and decision support workflow demonstrating the feasibility of achieving a real-time analysis and decision support system. We have prototyped automated analysis and optimization tools that facilitate a streamlined point-and-click process of refining molecular structures from experiment data, and we have demonstrated the feasibility enabling access to key facility resources as well as incorporating other community software tools for an enhanced end-user experience. We will continue to aggressively build on the existing prototype an incorporate several additional components and improvements that we have identified. Automated data analysis and optimization capabilities including detector boundary peak salvage, 1D Time of Flight peak fitting, and new peak integration methods and models are a priority. We will enhance the decision support capabilities of the user interface, and adapt GADDSS to support other instruments at the Spallation Neutron Source and High Flux Isotope Reactor. Commercial Applications and Other Benefits: The successful development of the proposed Department of Energy GADDSS project will directly benefit US neutron facilities. The early adopters, the TOPAZ, SNAP, MANDI, and IMAGINE instruments at the DOE BES facilities, will have the potential to more efficiently utilize beam experiment time. This will lead to more broad based application on other instruments and at other facilities worldwide.