ROYAL STATISTICAL SOCIETY
MANCHESTER LOCAL GROUP
A joint meeting with the Manchester Biostatistics group on the theme
Analysis Challenges arising from Post-Genomic high-Dimensional Datasets
Wednesday 8th October 2008 at MANDEC (Manchester Dental Education
Centre) Higher Cambridge Street, Manchester. 2-5 pm.
DAVID HOYLE (University of Manchester) ?Properties of sample
covariance matrices from modern high-dimensional genomic datasets ?
Modern biotechnology is providing a wealth of new data sets that are
high-dimensional but comparatively low sample size. This presents
challenges for traditional methods of analysis. Understanding the
properties of sample covariance matrices under such conditions has
become a recent research focus. In this talk I will outline new
findings and my own research in this area.
XIAYI KE (University of Manchester) ?Genome-wide association studies
of rheumatoid arthritis?
Genome wide association studies have been very successful in
identification of disease susceptibility loci for various complex
human diseases in recent 2-3 years. The Wellcome Trust Case Control
Consortium (WTCCC) is the best example of this development, where
seven common complex diseases, including rheumatoid arthritis (RA),
have been investigated and a number of disease susceptibility loci
have been uncovered. Before genome-wide association was applied to RA,
there were only two genetic loci known to be associated with RA. Now
more than ten novel loci have been identified. Despite these
successes, all these genetic factors identified so far can only
explain about 10% of the disease susceptibility variance, leaving wide
open where is the rest of genetic contribution and how to find each
element. In this talk, RA will also be used as an example to
illustrate the classical relationships between sample size, power,
effect size, and type I errors. Challenges in localization of causal
variants and detection of gene-gene interactions will also be discussed.
NEIL LAWRENCE (University of Manchester) ?Inference in Ordinary
Differential Equations with Latent Functions through Gaussian Processes?
In biochemical interaction networks is a key problem in estimation if
the structure and parameters of the genetic, metabolic and protein
interaction networks that underpin all biological processes. We
present a framework for Bayesian marginalisation of these latent
chemical species through Gaussian process priors. We demonstrate our
general approach on three different biological examples of single
input motifs, including both activation and repression of
transcription. We focus in particular on the problem of inferring
transcription factor activity when the concentration of active protein
cannot easily be measured. The uncertainty in the inferred
transcription factor activity can be integrated out in order to derive
a likelihood function that can be used for the estimation of
regulatory model parameters.
Tea will be served about mid-afternoon.
ALL ARE MOST WELCOME TO ATTEND! Please e-mail
[log in to unmask] if you will be attending
--
MHS-Medstats
|