So it begins! After months of preparation, our exhibition is finally ready for the crowd that will soon start to pour in at the @ArsElectronica festival. Don’t miss it if you’re around. And you should be around, the festival is incredibly cool!

All credits go to the amazing @ChrisHuetter who made it all happen. Originally trained in architecture, she dared to embark on an entirely different adventure with us on this project. This is her explaining it herself: youtu.be/tV3v4pp2YBc (10/n)

Our method can be used both to explore unknown network characteristics, as well as to accurately depict known ones. For example, I struggled for years to visually depict disease modules in the interactome. This can finally be done systematically! (8/n)

All credits go to the amazing @ChrisHuetter who made it all happen. Originally trained in architecture, she dared to embark on an entirely different adventure with us on this project. This is her explaining it herself: youtu.be/tV3v4pp2YBc (10/n)

We propose a series of layouts emphasizing various network characteristics that can be structural (e.g. node centrality), but also functional. One can choose pretty much anything. And since we know exactly what the node positions mean, we can finally interpret the layouts. (7/n)

We use a pretty simple trick allowing us to directly encode any quantity we like into node positions: dimensionality reduction! There are widely used algorithms for this, they are fast and can map anything onto various 2D and 3D geometries. (6/n)

#1 Size. The space on a computer screen or a piece of paper is just very limited. Displaying large networks will always be tough there. Not so in Virtual Reality, where you can blow a network up to the size of a cathedral and be right in the middle. rdcu.be/cHHRk (4/n)

I am particularly fond of this work, as it offers a solution to a problem that has been bugging me since my PhD: The hairball! Mark Newman put it perfectly in his classic review: (2/n)

All credits go to the amazing @izebuphamalai who performed this extensive tour de force of crunching dozens of datasets, extracting millions of interactions, thousands of diseases … This is what we looked like when we finally submitted the paper (10/n)

We found that rare disease genes are not scattered randomly across the multi-scale molecular network, but aggregate in phenotype specific disease modules. This finding forms the empirical basis for applying network medicine methodologies to rare diseases. (6/n)