Network Biology

Studying the molecular mechanisms in biological processes - tools and resources for pathway and network analysis

Pathway diagrams are everywhere: in textbooks, articles, posters and so on. Their utility to biologists as conceptual models is obvious. When modeled properly, they have also become immensely useful for computational analysis and interpretation of large-scale experimental data. In this talk, I will discuss key aspects in molecular network analysis related to data curation, data analysis, data integration and data visualization. I will highlight several pathway and network analysis related tools and resource developed at Maastricht University (including WikiPathways, PathViso and several Cytoscape apps) and their newest developments and features.

I will demonstrate the tools and their seamless integration into an analysis workflow to study the transcriptomics data of the diabetic liver.

Interpretation of genetic variations using network analysis

Single nucleotide polymorphisms (SNPs) from genome-wide association studies (GWAS) are challenging to interpret in the context of complex diseases. In particular, non-coding SNPs, which are the preponderant output of GWAS, are not systematically described in detailed analysis of biological impact. We addressed the challenge by integrating and visualizing different types of data in networks to build tissue specific genetic reference networks of SNPs associated with obesity. These networks are maps of SNPs, genes and pathways combined with other experimental data such as epigenetics and expression Quantitative Trait Loci (eQTLs). Stakeholders interested in obesity and personalized treatment can use the networks in various different ways. Experts in the field of obesity can explore the networks to generate novel hypothesis or confirm results related to the functional role of Body Mass Index SNPs and their possible effect on gene regulation by influencing epigenetic marks.

Phylogenetic inference of trees and networks - looking forwards!

At a high level, phylogenetics concerns the inference of plausible evolutionary histories (trees or, more generally, networks) given only measurements obtained from present-day data, typically (but by no means exclusively) aligned DNA or amino acid data. The fact that the tree/network is not given, but has to be found, gives phylogenetic inference a different flavor to many other areas of network biology where the network is already known -- and, computationally, it is extremely difficult! In this talk, I will first give a brief overview of classical and newer techniques for the inference of phylogenetic trees and networks, pointing out a number of commonly made mistakes in interpreting the meaning of such trees and networks. In the second part of the talk, I will (also briefly...) dispel the myth that phylogenetic inference is a backward-looking, genealogical stamp-collecting exercise. I will do this by demonstrating its use in comparative genomics as a tool for functional annotation (entities that have a recent common ancestry are likely to have certain shared functionality) and, in the biomedical domain, the growing use of phylogenetics in understanding tumor evolution and the spread of HIV.