Leukemia: Class Discrimination between AML and ALL using Microarrays

The objective of the experiment was to determine if the two types of leukemia could be discriminated by the different patterns of gene expression on a microarray. Machine Learning technology demonstrated the ability to discriminate between and diagnose two types of leukemia, Acute Myeloid Leukemia and Acute Lymphoblastic Leukemia, that require drastically differing courses of treatment but are difficult to distinguish medically. Analyzing a publicly available dataset (Golub, et al 1999) the RFE-SVM made an errorless distinction between the AML and ALL variants using a two-gene discriminator compared to a 3% error rate by the original authors. BIOwulf has exclusive rights to a patent application on this diagnostic discovery.