Name: Applications of Network Theory: From Mechanisms to Large-Scale Structure
Start: 2011-03-28T08:00:00+02:00
End: 2011-04-20T18:00:00+02:00
Location: No location set

Monday, March 28, 2011

1:20 PM Opening and orientation
Opening and orientation
1:20 PM - 1:30 PM
1:30 PM Respondent-driven Sampling on Directed Networks - Xin Lü (Karolinska Institute)
Respondent-driven Sampling on Directed Networks
- Xin Lü (Karolinska Institute)
1:30 PM - 2:30 PM
Respondent-driven sampling (RDS) circumvents the difficulties in sampling hard-to-reach population by using their social networks. It has been shown that the RDS is an effective method for generating handful sample. What’s more, under certain assumptions, unbiased estimates for population traits can be generated by weighting the sample with respondents’ personal network sizes. For these advantages the RDS has been widely used in HIV-related studies among high risk populations in recent years. However, despite the quite acknowledged evidences that most social networks are directed, all existing RDS estimators are based on assumption that the social relationships between respondents are reciprocal and consequently the RDS recruitment process happens on undirected networks. To investigate the potential bias brought by network directedness, and further generalize the RDS method, we propose several estimators that work both on directed networks and undirected networks. Performances of estimators are compared by RDS simulations on networks with different degrees of directedness, assortativity, indegree-outdegree correlation and homophily. Results reveal that the most robust estimator is the one which assumes the amount of individuals with trait A in the sample is proportional to the total indegree for individuals of group A in the population. A sensitivity test method is proposed when the sample indegree is not known. Given the widely existence of irreciprocal relationships among social society, we suggest the new estimator to be used in future RDS studies.
Tuesday, March 29, 2011

9:00 AM Breakfast at Nordita
Breakfast at Nordita
9:00 AM - 9:45 AM
10:00 AM Structural difference between public and private communication in an online community - Fariba Karimi (IceLab, Umeå University)
Structural difference between public and private communication in an online community
- Fariba Karimi (IceLab, Umeå University)
10:00 AM - 10:30 AM
We investigate an online community where there are two modes of communication. Either a user can reply to others in a public forum in such a way that we see who comments on whom; or they can send e-mail-like direct messages. In this data we investigate network structures (such as degree-distributions and assortativity), temporal structures such as response-, interevent times and activity levels. Furthermore, we measure combined structures from the different communication channels relating to structural-balance theory. Among other things, we find that in private communication, people keep feeling obliged to reply longer than in public discussions. We also observe a weak anti-correlation between activity levels in public and private communication respectively, suggesting that different personality types drive the large-scale structural evolution. We relate our findings to theories of social organization and human dynamics.
1:30 PM Network aspects of chromosome interactomes - Erik Aurell (KTH)
Network aspects of chromosome interactomes
- Erik Aurell (KTH)
1:30 PM - 2:30 PM
DNA is folded into increasingly complex yet highly mobile structures to organize the chromosomes. In this talk I will describe work with Rolf Ohlsson’s lab at KI where we have tried to quantify whether chromosome-chromosome interactions are random or not, and (if it is not) can we say something about the network structure of such interactions. The key experimental technique is high throughput sequencing of both read (short segements) known to interact with a known bait sequence, as well as chimeric reads containing pieces from different locations, in addition to the bait.
Wednesday, March 30, 2011
9:00 AM Metabolic networks, information, a null model and evolution - Petter Minnhagen (Umeå University)
Metabolic networks, information, a null model and evolution
- Petter Minnhagen (Umeå University)
9:00 AM - 10:00 AM
A simple general null-model, the Ikea network, is described and its properties are investigated. This network is argued to be the appropriate network for a random assembling of links, such that both which link is attached to which and the time-order are all distinct random possibilities. From this point of view it is the opposite of preferential attachment. This network is discussed in the context of metabolic networks, where the metabolism of an organism is reduced to a network of substances. The striking agreement between the Ikea network and the metabolic networks implies that the null model catches the overall features. Using the network structure measures clustering and assortativity, a small difference is nevertheless identified and is argued to imply a possible evolutionary pressure. This difference is manifested in a slight difference in the degree distribution.
1:30 PM Why do metabolic networks look like they do? The last decade and the next - Petter Holme (Coputational Biology)
Why do metabolic networks look like they do? The last decade and the next
- Petter Holme (Coputational Biology)
1:30 PM - 2:30 PM
This year is the 10th anniversary of the discovery that metabolic networks are scale-free. I will make a brief review of this decade of research relating network topology and function in metabolic reaction systems with a focus on our contributions. I discuss the hypothesis that network clusters correspond to functional modules. Metabolic network, however represented, are not as distinctly modular as the cartoon picture of intricately wired subsystems with few I/O-terminals. Does this reflect a trade-off between functionality and robustness, or is it an inevitable consequence of non-enzymatic reaction kinetics, or something else? I also discuss optimal levels of representations—if one uses a multiplex, directed, and perhaps bipartite, representation one can encode more information, but standard methods are harder to apply. If one goes for a simple-graph representation with vertices connected by undirected edges, then how can one encode as much functional information as possible? I will also mention how one can use other types of reaction systems, like reactions in planetary atmospheres, as null-models of metabolic networks. Finally I look forward and discuss open questions within reach with current and future data sets.
Thursday, March 31, 2011
9:00 AM Detecting community structures in uncertain networks using ensemble clustering - Johan Dahlin (Swedish Defense Research Agency)
Detecting community structures in uncertain networks using ensemble clustering
- Johan Dahlin (Swedish Defense Research Agency)
9:00 AM - 9:45 AM
During recent years many different methods to detect community structures in complex networks have been developed. Despite significant efforts from many different scientists from different fields, no completely satisfactory method to detect communities has been developed. In this talk, we present a method to merge the results from several different community detections run into a single final estimated community structure. Each run can either use a different method or else an ensemble of results from the same algorithm an be merged. We propose three different methods for the merge. Two of these are related to ensemble clustering methods used in standard data clustering problems. We apply these three methods on some different problems to demonstrate their usefulness. Some existing methods are stochastic in nature and generate different results for each run. Using the merging methods, one can use the collective information from the entire ensemble of possible community structures to find the most likely structure. This is shown to improve the performance of some stochastic algorithms. Another problem is related to the problem of uncertain Social Networks. In these networks, edges are not known with certainty to exist. Instead, a probability (or probability interval) is given for their existence. Using methods from statistical simulation, one can create an ensemble of networks that are consistent with the uncertain network. Applying existing methods for detecting community structures on each network from the ensemble creates a large number of different candidate community structures (one for each realization of the uncertain network). Applying the merging methods presented in this talk, one can merge (fuse) these different candidate structures into one, thus finding the community structure of the uncertain network.
Friday, April 1, 2011
9:00 AM Some results on hierarchical structures - Seung Ki Baek
Some results on hierarchical structures
- Seung Ki Baek
9:00 AM - 10:00 AM
Hierarchical structures have been considered as a way to study networks in a regular fashion. Our interest is especially in those obtained by tiling a hyperbolic plane, and we introduce numerical and analytical results about critical phenomena on these structures. We also show that the results can be extended to study phase transitions on the two-dimensional plane as well, including the percolation phenomena and the critical Potts model.
1:30 PM Exploring spatial networks with greedy navigators - Sang Hoon Lee (IceLab, Umeå University)
Exploring spatial networks with greedy navigators
- Sang Hoon Lee (IceLab, Umeå University)
1:30 PM - 2:30 PM
During the last decade, network research has focused on the global structural properties. Fewer studies take the local perspective of agents traveling the network. In this talk I will present a method that uses such a local perspective to integrate topological and spatial properties. This approach, we argue, will be even more important in this era of GPS-equipped smartphones, which give users ability to access local geometric information and navigate efficiently. We use a simple greedy spatial navigation strategy as a probe to explore spatial networks. These greedy navigators use geometric information to guide their moves and have a memory of their route in the network. We apply our method to several real-world networks of roads and railways. The results suggest that centrality measures have to be modified to incorporate the navigators’ behavior. We also see that removing some edges may actually enhance the routing efficiency, which is reminiscent of Braess’s paradox (caused by the conflict between user- and global optima). Furthermore, we present the reverse problem of optimizing the spatial layout of networks themselves to enhance the performance of the greedy spatial navigation. We relate these results, to the positioning of facilities and even architectural design to facilitate the behavior of humans.
Saturday, April 2, 2011
Sunday, April 3, 2011