This chapter describes the design of a prototype knowledge-based system for crystal and molecular structure determination from diffraction data. This system enhances current methods for the determination and interpretation of protein structures by incorporating direct methods probabilistic strategies, experience accumulated in the crystallographic databases, and knowledge representation and reasoning techniques for machine imagery.

The protein folding problem has been attacked from many directions. One set of approaches tries to find out correlations between short subsequences of proteins and the structures they form, using empirical information from crystallographic databases. AI research has repeatedly demonstrated the importance of representation in making these kinds of inferences. In this chapter, we describe an attempt to find a good representation for protein substructure. Our goal is to represent protein structures in such a way that they can, on the hand, reflect the enormous complexity and variety of different protein structures, and yet on the other hand facilitate the identification of similar substructures across different proteins.

The use of computers in chemistry and biochemistry has been widespread for many years, with machines performing many complex numerical calculations, e.g.

This chapter will discuss one example of how AI techniques are being integrated with, and extending, existing molecular biology sequence analysis methods. AI ideas of complex representations, pattern recognition, search, and machine learning have been applied to the task of inferring and recognizing structural patterns associated with molecular function. We wish to construct such patterns, and to recognize them in unknown molecules, based on information inferred solely from protein primary (amino acid) sequences.

Our understanding of any process can be measured by the extent to which a simulation we create mimics the real behavior of that process. Deviations of a simulation indicate either limitations or errors in our knowledge. In addition, these observed differences often suggest verifiable experimental hypotheses to extend our knowledge. The biochemical approach to understanding biological processes is essentially one of simulation. A biochemist typically prepares a cell-free extract that can mediate a well-described physiological process. The extract is then fractionated to purify the components that catalyze individual reactions.

Human scientists actively seek out information that bears on questions they have decided to pursue. They design experiments, explore the implications of the knowledge they have, refine their questions and test alternative ideas. Although many discoveries are the result of unexpected observations, these surprises take place in the context of an explicit pursuit of knowledge. Viewing scientific discovery as a kind of motivated action raises some basic issues common to goal-directed behavior generally: Where do desires (to know) come from? What are the actions that can be taken (to discover)? What are the resources those actions consume, and how are they allocated? How are decisions about selecting and combining actions made?

How could one represent with a single prototype the class of figures which have an even number of disconnected parts?

How could one represent with a single prototype the class of figures which have an even number of disconnected parts?

It is completely analogous to a set of difference equations that prescribes the transformation of a set of numbers through time. Given enough information about an individual, a program could be written that would describe the symbolic behavior of that individual. Each individual would be described by a different program, and those aspects of human problemsolving that are not idiosyncratic would emerge as the common structure and content of the programs of many individuals. But is it possible to write programs that do the kinds of manipulation that humans do? Given a specific protocol, such as the one of Figure 1, is it possible to induct the program of the subject? How well does a program fit the data? The remainder of the report will be devoted to answering some of these questions by means of the single example already presented. We will consider only how GPS behaves on the first part of the problem, and we will compare it in detail with the subject's behavior as revealed in sider the protocol. This will shed considerable light on how far we can con programs as theories of human problem-solving.

He then solves the simpler problem and uses the information obtained in the solution of the simpler problem in the solution of the original complex problem.