Seminar
Christina Ramirez Kitchen
Statistics In the Clinic: Phenotype-Genotype Relationships In Infectious Disease
In many scientific studies, such as those found in genetics, one often has many more parameters than
observations, the so-called "big p, small n problem". One needs to make efficient use of the data
while keeping in mind that the end result should have clinical or scientific relevance. In this talk,
I describe novel methods for this problem when the data sets are very large and also when they are small.
I use examples from HIV infection including predicting antiretroviral drug resistance, viral fitness and
coreceptor utilization. I will also describe novel models of intra-host HIV evolution of these phenotypes.
I then show how these models can be useful in the clinical setting.
