Database literature review: A new tool for experimental biology

The purpose of this review was to prepare data summaries that strengthen the connections between new research and the accumulated knowledge of the literature. To this end, we entered published data into a microcomputer database and then used the programming capabilities of the database to standardize data and generate summaries. As an example, we summarized data from papers that used stereology, a form of quantitative morphology, to describe ultrastructural changes in liver hepatocytes. We used standardized data to develop a ''one big experiment model'' of the literature and ultrastructural phenotypes to summarize structural analogies. The review illustrates how a database can be used to standardize results, reduce experimental biases, recover data, and search for biological correlations and patterns. A notable finding was that the ultrastructural phenotypes appear to be unique for each experimental data point, suggesting the use of this new data type as a generic biologic marker. We conclude from the results that a database literature review is an effective way of summarizing the literature and of creating new types of data.