Knowledge-based Interpretation of Gene Expression Array Studies Lawrence Hunter The new era of high throughput molecular instrumentation is generating important biomedical data at a rapidly increasing rate. The analysis and interpretation of this data must transcend traditional approaches to statistical hypothesis testing, toward more general computational support for biomedical discovery. In this talk I will argue that knowledge-based approaches, ranging from graphical statistical models with informative priors, to rule-based inference, to knowledge-based information extraction from natural language are the best way to meet this key challenge of the 21st century. |
Graybill Conference |