We present a strategy for generating and analyzing comprehensive genetic-interaction maps, termed E-MAPs (epistatic miniarray profiles), comprising quantitative measures of aggravating or alleviating interactions between gene pairs. Crucial to the interpretation of E-MAPs is their high-density nature made possible by focusing on logically connected gene subsets and including essential genes. Described here is the analysis of an E-MAP of genes acting in the yeast early secretory pathway. Hierarchical clustering, together with novel analytical strategies and experimental verification, revealed or clarified the role of many proteins involved in extensively studied processes such as sphingolipid metabolism and retention of HDEL proteins. At a broader level, analysis of the E-MAP delineated pathway organization and components of physical complexes and illustrated the interconnection between the various secretory processes. Extension of this strategy to other logically connected gene subsets in yeast and higher eukaryotes should provide critical insights into the functional/organizational principles of biological systems.