(Scientific Reports 5, 17386, 2015) we show that, by combining communities found in different types of genetic networks, one can identify candidate driving genes in cancer more easily than from the communities of individual networks
In the paper
Quantifying randomness in real networks(Nature Communications 6, 8627, 2015) we show that many features of real networks are the same if we randomize them by keeping their degree distributions, degree correlations and clustering. We conclude that many statistical properties of networks are consequences of such local observables
The paper
Attention decay in science(Journal of Informetrics 9, 734-745, 2015) shows that the exponential growth in the number of scientific publications makes papers get forgotten faster and faster. However, when time is measured in terms of the number of published items, the decay of attention of papers is approximately the same, suggesting that scientists are able to process a finite number of items
In the article
Triadic closure as a basic generating mechanism of communities in complex networks(Physical Review E 90, 042806, 2014) we show that network communities naturally emerge through growing mechanisms in which there is a sizeable probability of generating triangles. We conclude that triadic closure is responsible both for high transitivity and for community structure and that to explain the latter no additional hypothesis or mechanism is necessary
The paper Community detection in networks: structural communities versus ground truth (Physical Review E 90, 062805, 2014) shows that metadata may or may not be reflected in the topological communities detected by most methods commonly used. So one has to be careful to use metadata to validate community detection algorithms
In the paper Reputation and impact in scientific careers(PNAS 111, 15316-15321, 2014) we measured for the first time the effect of author reputation on the impact of their scientific output. Reputation helps in the initial phase of the history of the paper, until a given number of citations is reached, after which the paper is sustained by its own value and is not affected anymore by the author reputation
The time lag between the year when a Nobel-winning discovery is made and the year when it is awarded is growing exponentially, which might lead to the "extinction" of worthy Nobel Laureates. Take a look at our correspondence in Nature: Growing time lag threatens Nobels(Nature 508, 186, 2014). Why is it taking longer and longer for discoveries to be recognized?
In the paper Author Impact Factor: tracking the dynamics of individual scientific impact (Scientific Reports 4, 4880, 2014) we suggest a dynamic indicator of impact for individual scientists, the author impact factor, which is the average number of citations given by papers published in year t to papers published in a fixed period right before year t (e. g. the 5 previous years). It is the natural extension of the journal impact factor, and catches the variations of impact of scholars throughout their careers