ISSAC: Identification of Structured Signatures and Classifiers

ISSAC is a multiclass classification tool that can learn molecular signatures from high throughput gene expression data. ISSAC constructs a framework for disease diagnosis: a tree-structured hierarchy of the disease phenotypes under consideration based on agglomerative clustering, and learns panels of binary classifiers corresponding to the nodes and edges of the diagnostic hierarchy. Node classifiers distinguish the phenotype or group of phenotypes apart from all other phenotypes, and edge classifiers break ties by distinguishing between the two phenotype groups associated with each branch point.

Data and Software File(s):
Folder containing the ISSAC scripts
Description of how to run ISSAC