Seminar Data Science for the Life Sciences
Hosted by Amsterdam AI (AAI) Meetup
Details
room C0.110
Theme: data science for life sciences
Spreakers:
-Gunnar Klau (CWI/VU)
-Bas Teusink (VU)
-Johan Westerhuis (UvA)
-Bart Westerman (VU Medical Center)
met Q&A
moderator: Sanne Abeln (VU)
Speaker: Gunnar Klau, CWI/VU
Active subnetwork modules
Integrative network analysis methods provide robust interpretations of high-throughput molecular profile measurements in the context of gene networks. Active subnetwork modules are sets of genes that are responsible for the observed phenotypic changes and which are connected in the
network. Given appropriate scores, identifying them is equivalent to finding
maximum weight connected subgraphs. We describe Heinz, an exact approach to this problem, and eXamine, a Cytoscape app to explore the resulting modules, by means of an application to discovering virally deregulated signaling. See also http://youtu.be/LFGKek_pgjw .
Gunnar Klau leads the Life Sciences group at CWI and is extraordinary professor at the Department of Computer Science of the VU. His research focuses on combinatorial optimization for computational molecular biology.
Speaker: Bas Teusink, VUProbing the full metabolic potential of genomes
Metabolic networks can be reconstructed from genomic information, using bioinformatics tools and domain knowledge. Tools for automated reconstruction have been developed and they improve as more and more models are being developed and annotated. Such metabolic networks link metabolic reactions to proteins to genes and as such are being used as to integrate omics-data. We have combined optimisation techniqueswith experimental data to probe the full metabolic protential of the network of Bordetella pertussis, the causative agent of Whooping cough. These analyses revealed an enormous versatility in growth substrates, which on the one hand is alarming and on the other hand allowed us to develop vaccin-production media with spectacular improvements (2.4 fold).
Bas Teusink is professor in systems biology and integrative bioinformatics at the Amsterdam Institute for Molecules, Medicines and Systems of the VU University Amsterdam. He is biochemist by training and has an interest in developing and applying modeling and quantitative experimental approaches to unravel the design principles of biological networks.
Speaker: Johan Westerhuis, UvANetwork-based data integration
With the rise of new sequencing methodologies, the production of genomics data increases rapidly, but the integration of different data sets and the interpretation of the results lags behind. Traditional data fusion methods ignore the biological relationships such as the gene-protein-reaction associations or pathway associations.In this project we will develop new data fusion methods that are able combine data with pathway information. The analysis of the data will then focus on how pathway activity changes as a function of experimental conditions. These changes are explored within pathways as well as between pathways. Vehicles of accomplishing this are penalized methods and Bayesian methods.
Johan Westerhuis is an Assistant Professor in the Biosystems Data Analysis group at the Swammerdam Institute for Life Sciences (UvA). He has a background in developing data analysis methods for data of complex biological systems.
Speaker: Dr. Bart Westerman, VU Medical Center
Glioblastoma therapies evaluated:
from integrated patient data analysis to personalized therapy prediction
Cancer diagnostics based on identification of genetic drivers enables personalized medicine. This concept is now revolutionizing cancer therapy. However, glioblastoma, the most malignant brain tumor, is a highly heterogeneous tumor that remains incurable despite these novel developments. Prognosis for patients with glioblastoma is currently hampered by the absence of molecular markers that predict a clinical response. Recurrence of tumor growth is therefore commonly observed. We aim to understand glioblastoma through a robust molecular monitoring methodology and to use this information to improve current therapy by predicting drug-sensitivity towards personalized targets. For this, we have developed a bioinformatic drug-combination prediction model that uses FDA approved small molecule drugs that have been shown to cross the blood brain barrier. Our project integrates a number of novel and promising technologies for glioblastoma patients. This will function as a foundation for establishing the necessary tools to bridge the current gap between glioblastoma molecular profiling and a lack of therapy response.
Dr. Bart Westerman is an Assistant Professor in the interdisciplinary Neuro-oncology Research Group (NRG) at the VU Medical Center, Amsterdam. He has a background in genetic screens and cancer heterogeneity and integrates these themes in his current approach to develop drug-prediction methodologies.
---
