Toxicogenomics
We have more than five years experience in the analysis of toxicogenomic data.
Predictive Models
Dr. Michael Elashoff developed the predictive modeling methodology for Gene Logic's toxicogenomics predictive system that is currently in use. Gene Logic's database included hundreds of compounds profiled in time/dose studies, using both invitro and invivo experiments. This work included:
- Data normalization optimzed for toxicogenomic data
- Gene selection algorithms
- Predictive signatures for toxicity and specific pathologies
- Prediction scores and methods to put those scores into a meaningful biological context
- Cross validation of models
- Automation of the system
Cross Platform Prediction
Models can incorporate data from multiple Affymetrix platforms as well as two color (cDNA, Agilent) and other one color (Codelink, Illumina) platforms. See our cross-platform page for more details.
Phenotypic Anchoring
An emerging area in toxicogenomics is the tying of specific gene expression changes to observable pathological / clinical measurements. The genomic change could take the form of a fold change for a certain gene, a significant dysregulation in a pathway, or a change in a predictive model score. Contact us for more information on analytic approaches in this area.
Regulatory
Elashoff Consulting has a background in both the science side and the regulatory side of toxicogenomics and clinical drug development. For more information, please take a look at our regulatory page.
