Javier Cabrera
Location: (Hill Center, Bush Campus, Piscataway,NJ)
Personal Research Web Page: http://www.rci.rutgers.edu/~cabrera/
Keywords: Bioinformatics, gene networks, gene ontology, gene expressions, biological processes, pattern recognition, DNA barcode.
Posted on: Monday, June 8th, 2009
Broad Research Area: Graphics / Visualization, Numerical/Scientific Computing / HPC / Data-Intensive Scalable Computing, Scientific/Medical Informatics
Research Interests:
My area of research is to develop computational methods that will impact the way scientists analyze data from biological and medical genomics experiments. These methods will allow the incorporation of Gene Ontology information into modeling of gene expression data, and facilitate the simultaneous identification of biological processes that are affected by a treatment or another intervention. I propose to incorporate the gene correlation structure into our gene expression model as well as the indicator of the gene group (obtained from the Gene Ontology). The proposed models should allow more realistic fits to real data. We propose to study Bayesian and non-Bayesian alternatives in order to estimate the model parameters. In addition we propose a methodology to help scientists to develop a gene signature (or pattern) that enables the identification of groups of patients related to a condition or disease. This signature is based on a novel gene weighting scheme applied to the standard Random Forest algorithm. I also propose to extend this weighting scheme to other existing classification and prediction methods such as linear discriminant analysis, support vector machines, LASO and others. An additional interesting and important proposed application of these methods is to the analysis of DNA barcodes.
Contact Information:
email obfuscated - click to reveal
(Phone) 732-4852537
(Fax) 732-4453428
Department of Statistics and Biostatistics
Hill Center, Bush , Rutgers U
110 Frelinghuysen Rd
Piscataway, NJ08854
