Chapter 5 – Case-based reasoning driven gene annotation

This chapter examines case-based reasoning (CBR), a form of analogical reasoning, as a means of managing predictions of sparsely represented features in a biological database. CBR has the desirable property that it closely models the approach often taken by biologists when reasoning about structure and function in a new, partially characterized system from examples of well-characterized systems. This method of reasoning is particularly powerful in biology because it reflects the underlying evolutionary history that ties all modern life forms together. CBR can thus be viewed as a formalization of one form of reasoning about biological systems. From a database perspective, CBR provides a convenient framework for exploring issues on maintaining an accurate and consistent view of data automatically. From a biology perspective, CBR is an appropriate framework for making heuristic predictions over sparse features and for building a composite view of the current state of knowledge.