University of Cambridge > > Statistics > Bayesian Inference of Transmission Fidelity Rates of DNA Methylation Patterns

Bayesian Inference of Transmission Fidelity Rates of DNA Methylation Patterns

Add to your list(s) Download to your calendar using vCal

If you have a question about this talk, please contact rbg24.

A central question in epigenetics is how faithfully DNA methylation patterns are transmitted over cell generations. Our aim here is to estimate the rates of the maintenance and de novo methylation events that have occurred during the transmission process and have given rise to observed methylation patterns. A recently-emerging type of epigenetic data consists of binary DNA methylation patterns on double-stranded sequences in individual cells, each sequence containing a parent-daughter pair. However, parent and daughter strand in each pair are unidentified, the data are technically difficult and time-consuming to collect and they contain measurement error. We model these data at multiple sites jointly, thus using the whole-strand information, and considerably reduce confounding between parameters. We also adopt a hierarchical structure that allows for variation in rates across sites without an explosion in the effective number of parameters. Our context-specific priors capture the expected stationarity, or near-stationarity, of the stochastic process that generated the data analyzed here. This expected stationarity is further shown to greatly increase the precision of the estimation. Applying our model to data collected at the FMR1 locus on the hypermethylated X chromosome in females, we gain the following novel insights impossible to obtain under other existing analyses: (i) not accounting for measurement error has a large impact on estimation, leading to underestimation of maintenance rates and overestimation of de novo rates; and (ii) there is suggestive evidence of de novo events occurring on both parent and daughter strands. This is joint work with Diane Genereux, Charles Laird and Matthew Stephens.

This talk is part of the Statistics series.

Tell a friend about this talk:

This talk is included in these lists:

Note that ex-directory lists are not shown.


© 2006-2019, University of Cambridge. Contact Us | Help and Documentation | Privacy and Publicity