Jeanine Houwing

Data analytics and statistics for health

My work is motivated by my ongoing collaborations with clinicians, biologists, psychologists and chemists. For development of biostatistical methods, it is essential to know how the data are measured and which questions need to be answered.

My main research theme is random effects modelling, which is the basis of my work in family data: study design, testing and modelling genetic factors, risk prediction.

My work on Glycomics data was the motivation for a granted FP7 Consortium MIMOmics, for which I am the coordinator. For analysis of multimodal datasets in population cohorts, we develop network methods and prediction models. We use techniques from biostatistics, bioinformatics, machine learning, and physics. It appears that omics data improve the predictive performance, but not sufficiently for practical use. Adding family health data might advance the performance further.

Research in life and social sciences is data-driven. Data – if correctly analysed – provide insight in processes underlying health and disease. Data are big, noisy and incomplete. And data-sampling is complex. Research in statistical methods for data analysis is relevant and exciting.

Prof. J.J. (Jeanine) Houwing-Duistermaat

Sitting on the data. Why?

Why not sharing our data with other scientists? To stimulate multidisciplinary research? To accelerate scientific output by allowing more eyes to look at the data? To have fun together? In a comment which was published last week in the New England Journal of Medicine...

No Statistics – No Precision Medicine

A screening program tailored to my risk profile. Only receiving a treatment from which I will benefit. That is what we all want. Can big data help to achieve this goal? Yes it can. The reason is that big data allow identification of homogeneous patient groups. Within...