Home Page for Frank Bretz
Biostatistics & Statistical Reporting
WSJ-27.1.005
Novartis Pharma AG
CH-4002 Basel
Switzerland
Tel: ++41 - 61 - 324 4064
Fax: ++41 - 61 - 324 3039
Email:
frank.bretz@novartis.com
|
|
Research Interests
My primary research interests include multiple comparison procedures,
dose response experiments and adaptive designs. These topics are frequently
encountered in practice and have applications in such such different areas
as medicine, pharmaceutics, biology, physics, etc. In the past I have also
been interested in multidimensional integration routines to compute
probabilities and critical values from multivariate distribution functions
as well as in the design and analysis of microarray experimets.
Publications (and some selected papers)
Software
- The daMA
package contains functions for the efficient design and analysis of
factorial two-colour microarray experiments.
- The
multcomp package in R allows the evaluation of multiple comparison problems in
general linear models. A short description of the package is given
here.
Related macros in SAS can be found here.
- Programs to calculate multivariate normal or t-probabilities
using the approach of Genz (1992, 1993) and Genz and Bretz (1999, 2002):
SAS/IML
program for multivariate t-probabilities.
Simple SAS/IML
program for multivariate t-probabilities.
SAS/IML
program for multivariate normal probabilities.
SAS application,
which at run time calls executable files for the calculation of
MVT probabilities.
There are also implementations available in
R
and GAUSS.
- Programs for calculating orthant normal probabilities:
For small dimensions (k < 12) and tridiagonal correlation matrix
click here.
For higher dimensions and/or more general correlation matrices
click here.
For an algorithm to calculate level probabilities
click here.
For estimating the minimum effective dose using the LRT of Bartholomew
click here.
- Programs to calculate multivariate normal or t-probabilities
using the approach of Somerville and Bretz (2001a, b):