Register for: WIMSIG2017

Details of talk

TitleExtracting different disease information from genomic data: the case of repeat expansion disorders
PresenterMelanie Bahlo (Walter and Eliza Hall Institute)
Author(s)Melanie Bahlo
SessionBiostatistics and Bioinformatics
Time14:30:00 2017-09-26
Abstract


Most laypeople conceptualise disease causing mutations as single base pair
(A,G,C or T) changes in the normal human genome, which then lead to a damaged or
nonexistent protein, causing the disease. However, there are many other types of
genetic aberrations that cause disease. One of these is the expansion, beyond a
normal threshold, of short repetitive sequences of DNA. These are known to cause
$>$ 20 human diseases, mostly neurological, and includes Huntington’s disease.
Human genomes can now be quickly and relatively cheaply sequenced ($\sim$ \$1600
AUD). This has led to huge advances in medical genomics, with genome sequencing
moving into the clinic to diagnose disorders.

Until recently the known disease causing repeat expansion loci were not thought
to be able to be assessed for evidence of expansions with standard genome
sequencing data, but we and two other groups have now shown this can be done.
Our method uses an aggregated T-test statistic, with p-values determined by
simulation. We show that this performs very well for different types of genome
sequencing, being able to detect most individuals with known expansions. Repeat
data is fascinating, being affected by ethnicity and technological effects, as
well as modifier genetic loci. The next aim is to go beyond known disease loci,
to detect novel loci, and to look at the behavior of these repeats in both human
and malaria population-based studies and human neurological disease cohorts and
to further improve our test statistic and software.

% \centerline{\tt http://www.biorxiv.org/content/early/2017/06/30/157792}