Seminar: Dr Sach Mukherjee

Statistical and Machine Learning Approaches for the Exploration of Disease Heterogeneity

Date:
6 January 2015 16:00 hrs. - 17:00 hrs.
Location:
Mertens B, route 123
Title:
Statistical and Machine Learning Approaches for the Exploration of Disease Heterogeneity
Speaker(s):

Dr Sach Mukherjee, Biostatistics | Cambridge Institute of Public Health

Host(s):

Prof. Dr Jelle Goeman

06-01-2015 16:00:0006-01-2015 17:00:00Europe/AmsterdamStatistical and Machine Learning Approaches for the Exploration of Disease Heterogeneity Mertens B, route 123Rimlsrimls@radboudumc.nl

Remarks / more information:

Mukherjee _Sach -LRes -107x 150[1]Human diseases show considerable biological and clinical heterogeneity. Such heterogeneity is central to personalised medicine approaches that seek to exploit patient-specific data to inform clinical decision making and public health efforts. However, heterogeneity leads to nontrivial challenges in the analysis and interpretation of data, for example due to the fact that associations between variables (such as molecular expression levels or clinical covariates), or associations between such variables and outcomes of interest, may themselves depend on disease subtype. Such heterogeneity has important implications for many widely-used analytic approaches, including, among others, those involving clustering, regression, dimensionality reduction and network/causal models. Furthermore, these consequences are greatly exacerbated by the high-dimensional nature of current and emerging biomedical data. I will describe our ongoing efforts to develop statistical methods that can be used to investigate biological heterogeneity in a scalable, robust yet truly multi-dimensional manner, focusing on two case studies, the first concerning signalling downstream of receptor tyrosine kinases and the second exploring molecular data from cancer patient samples.

 



<< back to all events