Knowing the structure, dynamics and ligand binding specificity of proteins and nucleic acids is essential to understanding the mechanisms of human disease and to the optimal design of molecules that can intervene therapeutically in disease processes. Nuclear Magnetic Resonance (NMR) Spectroscopy is one of the most versatile techniques for obtaining this molecular information. This tremendous value of NMR in biomedical science is only realized through the application of powerful computational tools. Our goal in this project is to continue the development of, and add powerful new features to, an integrated software application for the computational analysis of NMR data. Having this effective software, NMRFx, for NMR is especially critical as we realize that deep insight into macromolecular structure and function comes not from a single technique like NMR, but from the complementary information from various techniques including NMR, X-ray crystallography, and Cryo-electron microscopy. Scientists are no longer seeing these techniques as specialized techniques usable by only domain experts, but as a collection of techniques that can and should be applied together. Without this NMRFx software, researchers must otherwise use a variety of tools from different labs that have different programming languages, scripting tools, naming conventions, graphical interfaces, documentation styles etc. The NMRFx software that is the subject of this project integrates signal processing, data analysis, visualization and macromolecular structure calculations based on NMR data. In the current project this software will be enhanced with the addition of an integrated deep learning library that will allow new types of data analyses using these new tools that are making major advances in scientific research. The performance of the software will be dramatically enhanced by the addition of a software library that allows computations to be dispatched to high-performance hardware like GPUs (graphical processing units) and FPGAs (Field Programmable Gate Arrays). This performance enhancement will allow much faster computations and allow use of algorithms that were previously too slow to use in interactive analyses. We will also complete the implementation of the remaining features necessary for the full analysis of NMR data. We propose that by providing new and experienced users with an enhanced version of our software application that has a common installation protocol, interface style, data structures, and overall design, we can substantially lower barriers to the use of NMR. NMRFx is making the use of NMR techniques more accessible to a wider population of researchers. The software, including the addition of the deep learning and acceleration libraries, will increase the contributions of NMR in a broad range of biological research areas that impact on human health.