IGSP Faculty

With full genome scans, DNA microarrays, complex models in structural biology, and other high throughput experimental techniques, the dimension of the data measured is often vastly higher than the number of observations taken.

Joe Lucas, PhD is an Assistant Research Professor at the IGSP. He received Masters degrees in Mathematics and Computer Science from the University of Pennsylvania, and completed his Ph. D. in Statistics at Duke University in December for 2006.

After completing his studies, he worked as a post-doctoral fellow in Mike West's lab in the Duke Department of Statistical Science studying the statistical issues related to high-dimensional, low sample size data sets, with particular interest in gene expression arrays.

Research Statement

The very high dimensional data available in many fields, but particularly in Biology, are stretching the standard statistical techniques beyond their capability. With full genome scans, DNA microarrays, complex models in structural biology, and other high throughput experimental techniques, the dimension of the data measured is often vastly higher than the number of observations taken.

Additionally, combining these vast sources in meaningful ways is a tremendous challenge, but offers tremendous potential to improve understanding of cellular and structural biology. For this type of data, standard statistical techniques are often inadequate.

My research is focused on questions of model and variable selection, especially related to high-dimensional multivariate modeling problems arising in biostatistics and genomics. My work to date involves a blend of modeling innovation for gene expression and structural biology together with computational statistics for model fitting and evaluation.

Learn more about Dr. Lucas's research in GenomeLIFE

Recent Publications