Statistical Bioinformatics Seminar: Professor Jean Yang -- Single cell data integrative analysis

Hi All, 

For our next Statistical Bioinformatics Seminar on Monday, we will be hosting
Professor Jean Yang from the University of Sydney.  The seminars are held at 1:00 pm on
Mondays at the Charles Perkins Centre, Level 3 Large meeting room.  The format of the
talk is approximately 40 minutes plus discussion.  Further information can be found on
the website https://www.maths.usyd.edu.au/u/SemConf/StatisticalBioinformatics.html

Monday October 21st 2019 
1:00 PM Seminar 
Level 3 Large Meeting Room 
Charles Perkins Centre 

Title: Single cell data integrative analysis 

Abstract: Recent advances in large scale single cell transcriptome profiling have
greatly expanded cell-type specific characterisation of complex biological systems.  It
enables discovery of many heterogeneous cell-types and differences in cell-type
proportions often carry biological significance.  A critical first step towards
understanding such differences is the accurate identification of cell types.  from the
complex tissues and organs.  A common approach achieved this by unsupervised clustering
followed by manual annotation according to marker gene expression.  With the increasing
availability of large collections of scRNA-seq datasets generated from the same tissues,
organs, and biological systems, as well as the comprehensive human and mouse cell
atlases, we are now at the transition point where supervised classification may be
trained to accurately classify cell types.  In this talk, I will discuss a number of
approaches develop at Sydney to address methodological challenges associated with single
cell data.  We will discuss a novel single cell differential composition (scDC) approach
that performs differential cell-type composition analysis via bootstrap resampling.  We
will introduce a multiscale classification framework (scClassify) for single cell
classification on a cell type hierarchy and ensemble learning where scClassify
effectively annotates cells at different levels of the cell type hierarchy.  Finally,
for a given training dataset, scClassify implements a sample size estimation procedure
to determine the number of cells required for accurate cell type classification at a
given cell type hierarchy level.  

About the Speaker: Professor Jean Yang is an applied statistician with expertise in
statistical bioinformatics.  She was awarded the 2015 Moran Medal in statistics from the
Australian Academy of Science in recognition of her work on developing methods for
molecular data arising in cutting edge biomedical research.  Her research stands at the
interface between medicine and methodology development and has centered on the
development of methods and the application of statistics to problems in -omics and
biomedical research.  She has made contributions to the development of novel statistical
methodology and software for the design and analysis of high-throughput biotechnological
data including that from microarrays, mass spectrometry and next generation sequencing.
Recently, much of her focus is on integration of multiple biotechnologies with clinical
data to answer a variety of scientific questions.  This includes developing various
approaches and methodologies in statistical machine learning and network analysis.  As a
statistician who works in the bioinformatics area, she enjoys research in a
collaborative environment, working closely with scientific investigators from diverse
backgrounds.