Giuseppe Facchetti

Computing global structural balance in large-scale signed social networks

Structural balance theory affirms that signed social networks (i.e., graphs whose signed edges represent friendly/hostile interactions among individuals) tend to be organized so as to avoid conflictual situations, corresponding to cycles of negative parity. Using an algorithm for ground-state calculation in large-scale Ising spin glasses, in this paper we compute the global level of balance of very large online social networks and verify that currently available networks are indeed extremely balanced. This property is explainable in terms of the high degree of skewness of the sign distributions on the nodes of the graph. In particular, individuals linked by a large majority of negative edges create mostly “apparent disorder,” rather than true “frustration.”

Common dynamical features of sensory adaptation in photoreceptors and olfactory sensory neurons

Sensory systems adapt, i.e., they adjust their sensitivity to external stimuli according to the ambient level. In this paper we show that single cell electrophysiological responses of vertebrate olfactory receptors and of photoreceptors to different input protocols exhibit several common features related to adaptation, and that these features can be used to investigate the dynamical structure of the feedback regulation responsible for the adaptation. In particular, we point out that two different forms of adaptation can be observed, in response to steps and to pairs of pulses. These two forms of adaptation appear to be in a dynamical trade-off: the more adaptation to a step is close to perfect, the slower is the recovery in adaptation to pulse pairs and viceversa. Neither of the two forms is explained by the dynamical models currently used to describe adaptation, such as the integral feedback model.

Predicting and characterizing selective multiple drug treatments for metabolic diseases and cancer

BackgroundIn the field of drug discovery, assessing the potential of multidrug therapies isa difficult task because of the combinatorial complexity (both theoretical andexperimental) and because of the requirements on the selectivity of the therapy.To cope with this problem, we have developed a novel method for the systematic insilico investigation of synergistic effects of currently available drugs ongenome-scale metabolic networks.ResultsThe algorithm finds the optimal combination of drugs which guarantees theinhibition of an objective function, while minimizing the side effect on the othercellular processes. Two different applications are considered: finding drugsynergisms for human metabolic diseases (like diabetes, obesity and hypertension)and finding antitumoral drug combinations with minimal side effect on the normalhuman cell. The results we obtain are consistent with some of the availabletherapeutic indications and predict new multiple drug treatments. A clusteranalysis on all possible interactions among the currently available drugsindicates a limited variety on the metabolic targets for the approved drugs.ConclusionThe in silico prediction of drug synergisms can represent an important tool forthe repurposing of drugs in a realistic perspective which considers also theselectivity of the therapy. Moreover, for a more profitable exploitation ofdrug-drug interactions, we have shown that also experimental drugs which have adifferent mechanism of action can be reconsider as potential ingredients of newmulticompound therapeutic indications. Needless to say the clues provided by acomputational study like ours need in any case to be thoroughly evaluatedexperimentally.

The phototransduction machinery in the rod outer segment has a strong efficacy gradient

Significance Phototransduction is now considered to be a quite thoroughly understood phenomenon. It is well known that new discs are continuously generated at the base of the outer segments (OSs) and old discs are shed at their tip, but the rod OSs are considered a well-stirred compartment with minor inhomogeneities. To verify this assumption and to better understand the machinery within the OS, we developed a new methodology to deliver highly localized lights. We found that, as the light stimulus is moved from the OS base toward its tip, the amplitude of saturating and single photon responses decreased by 5–10 times. This gradient of efficacy is attributed to a progressive loss of phosphodiesterase. Therefore, OSs are highly inhomogeneous compartments. Rod photoreceptors consist of an outer segment (OS) and an inner segment. Inside the OS a biochemical machinery transforms the rhodopsin photoisomerization into electrical signal. This machinery has been treated as and is thought to be homogenous with marginal inhomogeneities. To verify this assumption, we developed a methodology based on special tapered optical fibers (TOFs) to deliver highly localized light stimulations. By using these TOFs, specific regions of the rod OS could be stimulated with spots of light highly confined in space. As the TOF is moved from the OS base toward its tip, the amplitude of saturating and single photon responses decreases, demonstrating that the efficacy of the transduction machinery is not uniform and is 5–10 times higher at the base than at the tip. This gradient of efficacy of the transduction machinery is attributed to a progressive depletion of the phosphodiesterase along the rod OS. Moreover we demonstrate that, using restricted spots of light, the duration of the photoresponse along the OS does not increase linearly with the light intensity as with diffuse light.

Randomness in self-organized phenomena. A case study: Retinal angiogenesis

This note presents a review of recent work by the authors on angiogenesis, as a case study for analyzing the role of randomness in the formation of biological patterns. The mathematical description of the formation of new vessels is presented, based on a system of stochastic differential equations, coupled with a branching process, both of them driven by a set of relevant chemotactic underlying fields. A discussion follows about the possible reduction of complexity of the above approach, by mean field approximations of the underlying fields. The crucial role of randomness at the microscale is observed in order to obtain nontrivial realistic vessel networks.

A simple strategy guides the complex metabolic regulation in Escherichia coli

A way to decipher the complexity of the cellular metabolism is to study the effect of different external perturbations. Through an analysis over a sufficiently large set of gene knockouts and growing conditions, one aims to find a unifying principle that governs the metabolic regulation. For instance, it is known that the cessation of the microorganism proliferation after a gene deletion is only transient. However, we do not know the guiding principle that determines the partial or complete recovery of the growth rate, the corresponding redistribution of the metabolic fluxes and the possible different phenotypes. In spite of this large variety in the observed metabolic adjustments, we show that responses of E. coli to several different perturbations can always be derived from a sequence of greedy and myopic resilencings. This simple mechanism provides a detailed explanation for the experimental dynamics both at cellular (proliferation rate) and molecular level (13C-determined fluxes), also in case of appearance of multiple phenotypes. As additional support, we identified an example of a simple network motif that is capable of implementing this myopic greediness in the regulation of the metabolism.

Metabolic Adaptation Processes That Converge to Optimal Biomass Flux Distributions.

In simple organisms like E.coli, the metabolic response to an external perturbation passes through a transient phase in which the activation of a number of latent pathways can guarantee survival at the expenses of growth. Growth is gradually recovered as the organism adapts to the new condition. This adaptation can be modeled as a process of repeated metabolic adjustments obtained through the resilencings of the non-essential metabolic reactions, using growth rate as selection probability for the phenotypes obtained. The resulting metabolic adaptation process tends naturally to steer the metabolic fluxes towards high growth phenotypes. Quite remarkably, when applied to the central carbon metabolism of E.coli, it follows that nearly all flux distributions converge to the flux vector representing optimal growth, i.e., the solution of the biomass optimization problem turns out to be the dominant attractor of the metabolic adaptation process.

Partial inhibition and bilevel optimization in flux balance analysis

MotivationWithin Flux Balance Analysis, the investigation of complex subtasks, such as finding the optimal perturbation of the network or finding an optimal combination of drugs, often requires to set up a bilevel optimization problem. In order to keep the linearity and convexity of these nested optimization problems, an ON/OFF description of the effect of the perturbation (i.e. Boolean variable) is normally used. This restriction may not be realistic when one wants, for instance, to describe the partial inhibition of a reaction induced by a drug.ResultsIn this paper we present a formulation of the bilevel optimization which overcomes the oversimplified ON/OFF modeling while preserving the linear nature of the problem. A case study is considered: the search of the best multi-drug treatment which modulates an objective reaction and has the minimal perturbation on the whole network. The drug inhibition is described and modulated through a convex combination of a fixed number of Boolean variables. The results obtained from the application of the algorithm to the core metabolism of E.coli highlight the possibility of finding a broader spectrum of drug combinations compared to a simple ON/OFF modeling.ConclusionsThe method we have presented is capable of treating partial inhibition inside a bilevel optimization, without loosing the linearity property, and with reasonable computational performances also on large metabolic networks. The more fine-graded representation of the perturbation allows to enlarge the repertoire of synergistic combination of drugs for tasks such as selective perturbation of cellular metabolism. This may encourage the use of the approach also for other cases in which a more realistic modeling is required.

A rate-distortion theory for gene regulatory networks and its application to logic gate consistency

MOTIVATION A gene regulatory network in which the modes (activation/inhibition) of the transcriptional regulations are known and in which gene expression assumes boolean values can be treated as a system of linear equations over a binary field, i.e. as a constraint satisfaction problem for an information code. RESULTS For currently available gene networks, we show in this article that the distortion associated with the corresponding information code is much lower than expected from null models, and that it is close to (when not lower than) the Shannon bound determined by the rate-distortion theorem. This corresponds to saying that the distribution of regulatory modes is highly atypical in the networks, and that this atypicality greatly helps in avoiding contradictory transcriptional actions. Choosing a boolean formalism to represent the gene networks, we also show how to formulate criteria for the selection of gates that maximize the compatibility with the empirical information available on the transcriptional regulatory modes. Proceeding in this way, we obtain in particular that non-canalizing gates are upper-bounded by the distortion, and hence that the boolean gene networks are more canalizing than expected from null models.