Modelling heterogeneity in gene expression using the matrix-variate normal distribution
Add to your list(s)
Download to your calendar using vCal
If you have a question about this talk, please contact Dr Jack Bowden.
In this talk, we consider the problem of modeling gene expression levels
when measurements are made on multiple samples taken from the same
individual. The primary biological goal is to assess the heterogeneity
of the expression levels within each individual and, subsequently,
between all individuals. This type of data may arise in cancer studies
with multiple subsamples taken from each tumor. From a statistical
perspective our primary concern is the accurate estimation of the
association structure of the correlated samples and of the genes. For this purpose, we apply the matrix-variate normal distribution, which allows us to estimate separately the covariance/correlation matrix among the correlated subsamples and the genes. We derive covariance estimators and discuss their properties, and we develop test statistics for the sphericity and identity hypothesis testing of the covariance matrices. Finally, we illustrate the above using a real data set from a cancer study. This is a joint work with John Marioni and Simon Tavaré.
This talk is part of the MRC Biostatistics Unit Seminars series.
This talk is included in these lists:
Note that ex-directory lists are not shown.
|