Garrett Dancik
Graduate Student
Bioinformatics and Computational Biology
Department of Statistics
Rm 533, Science II

Courses

Stat430 - Fall 2007
R Tutorials [ BBSI 2008 || BCB544 ]
Stat104 – Fall 2006 / Spring 2007
Stat101 Lab – Spring 2006

Research

My general research interests are the following:

Computer modeling of the immune response
Statistical methods for the analysis of computer models
Calibration and Validation of computer and mathematical models

I have previously developed an agent-based model of Leishmania major infection which has been used to identify parameters that facilitate host suscpetibility to L. amazonensis . I am currently using a computer modeling approach to identify under what conditions mutations in the Rev protein result (or do not result) in a fitness advantage in EIAV infection.

I have also developed and maintain the R package mlegp , a Gaussian process (GP) modeling package designed primarily for the analysis of computer models. GPs are also commonly used for computer model calibration and validation. The package mlegp can fit GPs to heteroscedastic responses when the variance can be specified up to a multiplicative constant, and this is a first step in more flexible GP modeling of heteroscedastic responses.

Software

mlegp: an R package for maximum likelihood estimates of Gaussian processes, with diagnostic plots and sensitivity analysis