The long-term goal of this research is to provide significant speed-ups in understanding proteins through development of novel physics-based models in informatics. Proteins are involved in a large array of biological functions and their association with numerous diseases and disorders makes the timely understanding of their structure a subject with significant relevance to human health. As a step toward understanding proteins, the development of a broad inventory of protein structures and their rapid analysis is designated as a critical goal of structural biology. Experimental methods will continue to play a critical role, as a large number of novel protein folds remain unexplored. NMR spectroscopy is a key experimental tool in analyzing protein structures and a strong need for streamlining and extending its reach exists. The primary research focus of this work is the investigation and advancement of tools that will lead to significant speedups in understanding protein structures by building a probabilistic framework that integrates informatics and physical models. The strategy is to combine the use of informatics tools and physical modeling that is needed in order to rapidly evaluate, merge multiple data sources, and facilitate efficient building and analysis of protein inventories. The proposed approach has already lead to innovative tools that have demonstrated quantifiable advances in the practice of NMR structure determination. The applicant has three research goals during this grant period: 1) investigate approaches to combining present tools developed by the PI into a complete paradigm with the aim of addressing fast and robust structure determination of small to moderate size proteins, 2) investigate distillation and extension of methods into a set of core tools that could form the basis for novel tools for rapid determination of protein folds and structure of larger proteins, and 3) take exploratory steps toward understanding the question of "how much each new tool contributes to our understanding of protein space." The basic idea behind these methods is to devise a family of physical models and use informatics tools to find the most 'successful' model. These tools will facilitate production of timely information regarding proteins' functions by speeding up and streamlining the use of experimental data in structure determination. The accelerated progress toward understanding proteins will have a direct and significant impact on advancing the safeguards of human health.