ROYAL STATISTICAL SOCIETY MEDICAL SECTION
Tuesday 29 November 2005 2.00-5.00 (tea at 3.10).
To be held at the RSS, 12 Errol Street, London EC1Y 8LX
(directions http://www.rss.org.uk/about/direction.html)
Pre-registration is recommended - please email [log in to unmask] <mailto:[log in to unmask]> <mailto:[log in to unmask]> or tel. 020 7638 8998 to register.
‘Statistics in the post-genomic age’
A half-day meeting of presentations
2.00 Wally Gilks. MRC Biostatistics Unit, Cambridge.
2.35 Heather J. Cordell, Department of Medical Genetics, University of
Cambridge.
3.10 Tea
3.40 Sylvia Richardson, Imperial College, London.
4.20 Jenny Barrett, Division of Genetic Epidemiology, University of Leeds.
5.00 Close
Titles and abstracts
Statistical Bioinformatics: an overview.
Wally Gilks. MRC Biostatistics Unit, Cambridge.
The rapid expansion of DNA, protein, gene expression and other genomic
databases has spawned the field of bioinformatics. Work in this field
spans the development of databases and algorithms, and research in basic
biology, genetics and medicine. I will briefly review the state of the
art, focussing on current and potential areas of involvement of
statisticians.
Regression methods for SNP data in case/control and family studies.
Heather J. Cordell, Department of Medical Genetics, University of Cambridge.
Here I discuss methods of design and analysis of genetic association
studies. There are many similarities between genetic association studies
and classical epidemiological studies of environmental risk factors, but
there are also issues specific to studies of genetic risk factors such
as the use of certain family-based designs, accounting for different
underlying genetic mechanisms and the impact of population history.
Statistical analysis of gene expression data
Sylvia Richardson, Imperial College, London.
The powerful technology of cDNA or oligonucleotide microarray makes it
possible to study simultaneously the expression of thousands of genes in
different samples. To interpret and model this vast body of data poses
interesting statistical challenges. Issues of signal extraction,
normalisation and how to find groups of genes that are differentially
expressed will be discussed. The flexibility and benefits of using a
Bayesian hierarchical modelling framework to perform such analyses will
be illustrated.
Statistical issues in clinical proteomics
Jenny Barrett, Division of Genetic Epidemiology, University of Leeds.
Proteomics concerns the identification and characterisation of expressed
proteins in an organism. The subject will be introduced, focusing on
SELDI mass spectrometry profiles. Statistical issues in experimental
design, pre-processing and peak detection will be discussed, before
moving on to identification of differences between peak profiles in
different sample groups. Methods will be illustrated with
clinically-motivated examples.
|