In the last several years, the focus on protein structure and folding has begun to shift from physical modeling to pattern matching. What has happened is that, against initial expectation, the structure of enough proteins has been learned that there are definite patterns that are observable and that have great predictive value both for structure and function. For instance, recently, some researchers have started exploring some subtasks in protein structure prediction from a discrete algorithmic point of view. Investigators have had some initial success on some problems of protein geometry, e.g., three dimensional surface matching, and in predicting, or detecting structure from sequence data, e.g., motif recognition from sequence data. Thus, the study of protein structure is becoming much more the study of sequence and critical conserved features from sequence, and that makes this topic very much a topic where discrete mathematicians can make contributions. These ideas are far from mature and their long-term relevance is unclear. Still, a mini-workshop or two will give us an inexpensive way to explore the area and interchange ideas.