OPTIMIZATION-BASED EXPLORATION OF PARKINSON'S DISEASE MICROARRAY EXPERIMENTS
Janice Garcia1, Trujillo Alto1, Clara Isaza2, Xavier Aguilar1.
1University of Puerto Rico, Mayaguez Campus, Mayaguez, PR, 2Ponce Health Sciences University, Ponce, PR.
Parkinson’s disease (PD) is a progressive disorder of the nervous system that affects movement. In order to study PD, it is possible to make use of high-throughput techniques to detect important biological signals. Such is the case for microarrays which quantify relative expression of tens of thousands of genes simultaneously. In our research group, a 2-step analysis pipeline has been proposed to detect highly differentially expressed genes and determine the most correlated path among them as a proxy to the potential signaling pathway. The methods incorporated in this pipeline are based on mathematical optimization which is deemed novel. These methods include multicriteria optimization and network flow optimization. In particular, in the latter, the traveling salesman problem and the minimum spanning tree mathematical formulations were used. This research focuses on generating knowledge about Parkinson’s disease in the form of a potential genetic signature and a potential signaling pathway. Preliminary results in potential biomarkers and signaling paths will be discussed as well as the role that this information could play in the understanding of the disease. The analysis of a single microarray experiment has been used to obtain biologically relevant information, and the results of the proposed analysis strategy have been already validated in previous biological experiments, supporting the potential and high discriminating power of this strategy.