Kathleen Marchal

Kathleen Marchal, Dept of Plant Biotechnology and Bioinformatics, Fac of Sciences; Department of Informatics, IDLab, IMEC.
(ORCID)

Graph-based approaches to improve cancer driver and drug target identification

Increasingly innovation is driven by the use of omics technologies, resulting in an accumulation of historical data in both the public and private domain. Optimally exploiting such historical data is key to the discovery of novel targets and /or biomarkers. Graph-based methods provide an ideal data-analytics framework to deal with historical data. They use an underlying network model in which heterogeneous data can be integrated in an intuitive way. In addition, they often steer their analysis by means of prior knowledge on molecular interactions. This use of prior information allows coping with statistical underdetermination (the insufficient number of independent samples) while at the same time unveiling molecular mechanisms. We will illustrate the potential of such network-based approaches for the analysis of historical data with two case studies. In a first study we present GoNetic, a versatile network-based framework for integrative network-based driver identification. A second study focuses on the use of link prediction to reprioritize dependencies derived from LOF screening data.

Short bio

Kathleen Marchal obtained a PhD in Molecular Microbiology in 1999. From 1999-2004 she performed a postdoc in bioinformatics in the department of electrical engineering (ESAT/SCD, KULeuven). In 2004 she became assistant and later associate professor at the KU Leuven. In 2011, she became associate professor in bioinformatics at Ghent University (Dept of Plant Biotechnology and Bioinformatics, Fac of Sciences). The interdisciplinary research group of K. Marchal is physically located at the Dept of Information Technology (Fac of Engineering, UGhent).

The main expertise of the group consists of developing computational methods for outstanding problems in systems genetics with an emphasis on network-based methods for data integration.