Frankfurt Institute for Advanced Studies (FIAS)
Refine
Year of publication
Document Type
- Preprint (962)
- Article (753)
- Conference Proceeding (27)
- Doctoral Thesis (18)
- Part of Periodical (6)
- Contribution to a Periodical (3)
- Part of a Book (2)
- Diploma Thesis (1)
- Master's Thesis (1)
- Review (1)
Has Fulltext
- yes (1774) (remove)
Is part of the Bibliography
- no (1774)
Keywords
- Heavy Ion Experiments (21)
- Hadron-Hadron Scattering (11)
- Hadron-Hadron scattering (experiments) (11)
- LHC (10)
- Heavy-ion collisions (8)
- Heavy-ion collision (7)
- heavy-ion collisions (7)
- schizophrenia (7)
- Black holes (6)
- Equation of state (5)
Institute
- Frankfurt Institute for Advanced Studies (FIAS) (1774)
- Physik (1315)
- Informatik (1008)
- Medizin (64)
- MPI für Hirnforschung (31)
- Ernst Strüngmann Institut (26)
- Biowissenschaften (22)
- Psychologie (13)
- Biochemie und Chemie (12)
- Helmholtz International Center for FAIR (7)
The detailed biophysical mechanisms through which transcranial magnetic stimulation (TMS) activates cortical circuits are still not fully understood. Here we present a multi-scale computational model to describe and explain the activation of different pyramidal cell types in motor cortex due to TMS. Our model determines precise electric fields based on an individual head model derived from magnetic resonance imaging and calculates how these electric fields activate morphologically detailed models of different neuron types. We predict neural activation patterns for different coil orientations consistent with experimental findings. Beyond this, our model allows us to calculate activation thresholds for individual neurons and precise initiation sites of individual action potentials on the neurons’ complex morphologies. Specifically, our model predicts that cortical layer 3 pyramidal neurons are generally easier to stimulate than layer 5 pyramidal neurons, thereby explaining the lower stimulation thresholds observed for I-waves compared to D-waves. It also shows differences in the regions of activated cortical layer 5 and layer 3 pyramidal cells depending on coil orientation. Finally, it predicts that under standard stimulation conditions, action potentials are mostly generated at the axon initial segment of cortical pyramidal cells, with a much less important activation site being the part of a layer 5 pyramidal cell axon where it crosses the boundary between grey matter and white matter. In conclusion, our computational model offers a detailed account of the mechanisms through which TMS activates different cortical pyramidal cell types, paving the way for more targeted application of TMS based on individual brain morphology in clinical and basic research settings.
Current theories of schizophrenia (ScZ) posit that the symptoms and cognitive dysfunctions arise from a dysconnection syndrome. However, studies that have examined this hypothesis with physiological data at realistic time scales are so far scarce. The current study employed a state-of-the-art approach using Magnetoencephalography (MEG) to test alterations in large-scale phase synchronization in a sample of n = 16 chronic ScZ patients, 10 males and n = 19 healthy participants, 10 males, during a perceptual closure task. We identified large-scale networks from source reconstructed MEG data using data-driven analyses of neuronal synchronization. Oscillation amplitudes and interareal phase-synchronization in the 3–120 Hz frequency range were estimated for 400 cortical parcels and correlated with clinical symptoms and neuropsychological scores. ScZ patients were characterized by a reduction in γ-band (30–120 Hz) oscillation amplitudes that was accompanied by a pronounced deficit in large-scale synchronization at γ-band frequencies. Synchronization was reduced within visual regions as well as between visual and frontal cortex and the reduction of synchronization correlated with elevated clinical disorganization. Accordingly, these data highlight that ScZ is associated with a profound disruption of transient synchronization, providing critical support for the notion that core aspect of the pathophysiology arises from an impairment in coordination of distributed neural activity.
The goal of heavy ion reactions at low beam energies is to explore the QCD phase diagram at high net baryon chemical potential. To relate experimental observations with a first order phase transition or a critical endpoint, dynamical approaches for the theoretical description have to be developed. In this summary of the corresponding plenary talk, the status of the dynamical modeling including the most recent advances is presented. The remaining challenges are highlighted and promising experimental measurements are pointed out.
The three-dimensional structure determination of RNAs by NMR spectroscopy relies on chemical shift assignment, which still constitutes a bottleneck. In order to develop more efficient assignment strategies, we analysed relationships between sequence and 1H and 13C chemical shifts. Statistics of resonances from regularly Watson– Crick base-paired RNA revealed highly characteristic chemical shift clusters. We developed two approaches using these statistics for chemical shift assignment of double-stranded RNA (dsRNA): a manual approach that yields starting points for resonance assignment and simplifies decision trees and an automated approach based on the recently introduced automated resonance assignment algorithm FLYA. Both strategies require only unlabeled RNAs and three 2D spectra for assigning the H2/C2, H5/C5, H6/C6, H8/C8 and H10/C10 chemical shifts. The manual approach proved to be efficient and robust when applied to the experimental data of RNAs with a size between 20 nt and 42 nt. The more advanced automated assignment approach was successfully applied to four stemloop RNAs and a 42 nt siRNA, assigning 92–100% of the resonances from dsRNA regions correctly. This is the first automated approach for chemical shift assignment of non-exchangeable protons of RNA and their corresponding 13C resonances, which provides an important step toward automated structure determination of RNAs.
We present results on transverse momentum (pT) and rapidity (y) differential production cross sections, mean transverse momentum and mean transverse momentum square of inclusive J/ψ and ψ(2S) at forward rapidity (2.5 < y < 4) as well as ψ(2S)-to-J/ψ cross section ratios. These quantities are measured in pp collisions at center of mass energies s√=5.02 and 13 TeV with the ALICE detector. Both charmonium states are reconstructed in the dimuon decay channel, using the muon spectrometer. A comprehensive comparison to inclusive charmonium cross sections measured at s√=2.76, 7 and 8 TeV is performed. A comparison to non-relativistic quantum chromodynamics and fixed-order next-to-leading logarithm calculations, which describe prompt and non-prompt charmonium production respectively, is also presented. A good description of the data is obtained over the full pT range, provided that both contributions are summed. In particular, it is found that for pT > 15 GeV/c the non-prompt contribution reaches up to 50% of the total charmonium yield.
The ability to learn sequential behaviors is a fundamental property of our brains. Yet a long stream of studies including recent experiments investigating motor sequence learning in adult human subjects have produced a number of puzzling and seemingly contradictory results. In particular, when subjects have to learn multiple action sequences, learning is sometimes impaired by proactive and retroactive interference effects. In other situations, however, learning is accelerated as reflected in facilitation and transfer effects. At present it is unclear what the underlying neural mechanism are that give rise to these diverse findings. Here we show that a recently developed recurrent neural network model readily reproduces this diverse set of findings. The self-organizing recurrent neural network (SORN) model is a network of recurrently connected threshold units that combines a simplified form of spike-timing dependent plasticity (STDP) with homeostatic plasticity mechanisms ensuring network stability, namely intrinsic plasticity (IP) and synaptic normalization (SN). When trained on sequence learning tasks modeled after recent experiments we find that it reproduces the full range of interference, facilitation, and transfer effects. We show how these effects are rooted in the network’s changing internal representation of the different sequences across learning and how they depend on an interaction of training schedule and task similarity. Furthermore, since learning in the model is based on fundamental neuronal plasticity mechanisms, the model reveals how these plasticity mechanisms are ultimately responsible for the network’s sequence learning abilities. In particular, we find that all three plasticity mechanisms are essential for the network to learn effective internal models of the different training sequences. This ability to form effective internal models is also the basis for the observed interference and facilitation effects. This suggests that STDP, IP, and SN may be the driving forces behind our ability to learn complex action sequences.
Two theories address the origin of repeating patterns, such as hair follicles, limb digits, and intestinal villi, during development. The Turing reaction–diffusion system posits that interacting diffusible signals produced by static cells first define a prepattern that then induces cell rearrangements to produce an anatomical structure. The second theory, that of mesenchymal self-organisation, proposes that mobile cells can form periodic patterns of cell aggregates directly, without reference to any prepattern. Early hair follicle development is characterised by the rapid appearance of periodic arrangements of altered gene expression in the epidermis and prominent clustering of the adjacent dermal mesenchymal cells. We assess the contributions and interplay between reaction–diffusion and mesenchymal self-organisation processes in hair follicle patterning, identifying a network of fibroblast growth factor (FGF), wingless-related integration site (WNT), and bone morphogenetic protein (BMP) signalling interactions capable of spontaneously producing a periodic pattern. Using time-lapse imaging, we find that mesenchymal cell condensation at hair follicles is locally directed by an epidermal prepattern. However, imposing this prepattern’s condition of high FGF and low BMP activity across the entire skin reveals a latent dermal capacity to undergo spatially patterned self-organisation in the absence of epithelial direction. This mesenchymal self-organisation relies on restricted transforming growth factor (TGF) β signalling, which serves to drive chemotactic mesenchymal patterning when reaction–diffusion patterning is suppressed, but, in normal conditions, facilitates cell movement to locally prepatterned sources of FGF. This work illustrates a hierarchy of periodic patterning modes operating in organogenesis.
Dendrites form predominantly binary trees that are exquisitely embedded in the networks of the brain. While neuronal computation is known to depend on the morphology of dendrites, their underlying topological blueprint remains unknown. Here, we used a centripetal branch ordering scheme originally developed to describe river networks—the Horton-Strahler order (SO)–to examine hierarchical relationships of branching statistics in reconstructed and model dendritic trees. We report on a number of universal topological relationships with SO that are true for all binary trees and distinguish those from SO-sorted metric measures that appear to be cell type-specific. The latter are therefore potential new candidates for categorising dendritic tree structures. Interestingly, we find a faithful correlation of branch diameters with centripetal branch orders, indicating a possible functional importance of SO for dendritic morphology and growth. Also, simulated local voltage responses to synaptic inputs are strongly correlated with SO. In summary, our study identifies important SO-dependent measures in dendritic morphology that are relevant for neural function while at the same time it describes other relationships that are universal for all dendrites.
I summarize recent developments in the hard-thermal-loop approach to QCD. I first discuss a finite-temperature and -density calculation of QCD thermodynamics at NNLO from the hard-thermal-loop perturbation theory. I then discuss a generalization of the hard-thermal-loop framework to the magnetic scale g2T, from which a novel non-Abelian massless mode is uncovered.
The pA system is typically regarded in heavy ion collisions as a “cold” nuclear matter environment and thought to isolate and identify initial state effects due to the presence of multiple nucleons in the incoming nucleus. Moreover, pA collisions bridge the gap between peripheral AA collisions and the pp baseline to create a more complete understanding of underlying production mechanisms and how they evolve with multiplicity. Recent measurements at both RHIC and the LHC provide an indication, however, that the “cold” nuclear matter picture may be somewhat naïve.
Recent LHC results from the 2013 p–Pb run at √sNN = 5.02 TeV will be discussed.
Overrepresentation of bidirectional connections in local cortical networks has been repeatedly reported and is a focus of the ongoing discussion of nonrandom connectivity. Here we show in a brief mathematical analysis that in a network in which connection probabilities are symmetric in pairs, Pij = Pji, the occurrences of bidirectional connections and nonrandom structures are inherently linked; an overabundance of reciprocally connected pairs emerges necessarily when some pairs of neurons are more likely to be connected than others. Our numerical results imply that such overrepresentation can also be sustained when connection probabilities are only approximately symmetric.
Criticality meets learning : criticality signatures in a self-organizing recurrent neural network
(2017)
Many experiments have suggested that the brain operates close to a critical state, based on signatures of criticality such as power-law distributed neuronal avalanches. In neural network models, criticality is a dynamical state that maximizes information processing capacities, e.g. sensitivity to input, dynamical range and storage capacity, which makes it a favorable candidate state for brain function. Although models that self-organize towards a critical state have been proposed, the relation between criticality signatures and learning is still unclear. Here, we investigate signatures of criticality in a self-organizing recurrent neural network (SORN). Investigating criticality in the SORN is of particular interest because it has not been developed to show criticality. Instead, the SORN has been shown to exhibit spatio-temporal pattern learning through a combination of neural plasticity mechanisms and it reproduces a number of biological findings on neural variability and the statistics and fluctuations of synaptic efficacies. We show that, after a transient, the SORN spontaneously self-organizes into a dynamical state that shows criticality signatures comparable to those found in experiments. The plasticity mechanisms are necessary to attain that dynamical state, but not to maintain it. Furthermore, onset of external input transiently changes the slope of the avalanche distributions – matching recent experimental findings. Interestingly, the membrane noise level necessary for the occurrence of the criticality signatures reduces the model’s performance in simple learning tasks. Overall, our work shows that the biologically inspired plasticity and homeostasis mechanisms responsible for the SORN’s spatio-temporal learning abilities can give rise to criticality signatures in its activity when driven by random input, but these break down under the structured input of short repeating sequences.
The detailed biophysical mechanisms through which transcranial magnetic stimulation (TMS) activates cortical circuits are still not fully understood. Here we present a multi-scale computational model to describe and explain the activation of different pyramidal cell types in motor cortex due to TMS. Our model determines precise electric fields based on an individual head model derived from magnetic resonance imaging and calculates how these electric fields activate morphologically detailed models of different neuron types. We predict neural activation patterns for different coil orientations consistent with experimental findings. Beyond this, our model allows us to calculate activation thresholds for individual neurons and precise initiation sites of individual action potentials on the neurons’ complex morphologies. Specifically, our model predicts that cortical layer 3 pyramidal neurons are generally easier to stimulate than layer 5 pyramidal neurons, thereby explaining the lower stimulation thresholds observed for I-waves compared to D-waves. It also shows differences in the regions of activated cortical layer 5 and layer 3 pyramidal cells depending on coil orientation. Finally, it predicts that under standard stimulation conditions, action potentials are mostly generated at the axon initial segment of cortical pyramidal cells, with a much less important activation site being the part of a layer 5 pyramidal cell axon where it crosses the boundary between grey matter and white matter. In conclusion, our computational model offers a detailed account of the mechanisms through which TMS activates different cortical pyramidal cell types, paving the way for more targeted application of TMS based on individual brain morphology in clinical and basic research settings.
BACKGROUND: The analysis of microarray time series promises a deeper insight into the dynamics of the cellular response following stimulation. A common observation in this type of data is that some genes respond with quick, transient dynamics, while other genes change their expression slowly over time. The existing methods for detecting significant expression dynamics often fail when the expression dynamics show a large heterogeneity. Moreover, these methods often cannot cope with irregular and sparse measurements.
RESULTS: The method proposed here is specifically designed for the analysis of perturbation responses. It combines different scores to capture fast and transient dynamics as well as slow expression changes, and performs well in the presence of low replicate numbers and irregular sampling times. The results are given in the form of tables including links to figures showing the expression dynamics of the respective transcript. These allow to quickly recognise the relevance of detection, to identify possible false positives and to discriminate early and late changes in gene expression. An extension of the method allows the analysis of the expression dynamics of functional groups of genes, providing a quick overview of the cellular response. The performance of this package was tested on microarray data derived from lung cancer cells stimulated with epidermal growth factor (EGF).
CONCLUSION: Here we describe a new, efficient method for the analysis of sparse and heterogeneous time course data with high detection sensitivity and transparency. It is implemented as R package TTCA (transcript time course analysis) and can be installed from the Comprehensive R Archive Network, CRAN. The source code is provided with the Additional file 1.
Study of hard core repulsive interactions in an hadronic gas from a comparison with lattice QCD
(2016)
We study the influence of hard-core repulsive interactions within the Hadron-Resonace Gas model in comparison to first principle calculation performed on a lattice. We check the effect of a bag-like parametrization for particle eigenvolume on flavor correlators, looking for an extension of the agreement with lattice simulations up to higher temperatures, as was yet pointed out in an analysis of hadron yields measured by the ALICE experiment. Hints for a flavor depending eigenvolume are present.
The future heavy-ion experiment CBM (FAIR/GSI, Darmstadt, Germany) will focus on measurement of very rare probes at interaction rates up to 10 MHz with data flow of up to 1 TB/s. The beam will provide free stream of beam particles without bunch structure. That requires full online event reconstruction and selection not only in space, but also in time, so-called 4D event building and selection.
The FLES (First-Level Event Selection) reconstruction and selection package consists of several modules: track finding, track fitting, short-lived particles finding, event building and event selection. A time-slice is reconstructed in parallel between cores within a same CPU, thus minimizing the communication between CPUs. After all tracks are found and fitted in 4D, they are collected into clusters of tracks originated from common primary vertices, which then are fitted, thus identifying 4D interaction points registered within the time-slice. Secondary tracks are associated with primary vertices according to their estimated production time. After that, short-lived particles are found and the full event building process is finished. The last stage of the FLES package is the selection of events according to the requested trigger signatures.
Neural oscillations at low- and high-frequency ranges are a fundamental feature of large-scale networks. Recent evidence has indicated that schizophrenia is associated with abnormal amplitude and synchrony of oscillatory activity, in particular, at high (beta/gamma) frequencies. These abnormalities are observed during task-related and spontaneous neuronal activity which may be important for understanding the pathophysiology of the syndrome. In this paper, we shall review the current evidence for impaired beta/gamma-band oscillations and their involvement in cognitive functions and certain symptoms of the disorder. In the first part, we will provide an update on neural oscillations during normal brain functions and discuss underlying mechanisms. This will be followed by a review of studies that have examined high-frequency oscillatory activity in schizophrenia and discuss evidence that relates abnormalities of oscillatory activity to disturbed excitatory/inhibitory (E/I) balance. Finally, we shall identify critical issues for future research in this area.
Introduction: Neuronal death and subsequent denervation of target areas are hallmarks of many neurological disorders. Denervated neurons lose part of their dendritic tree, and are considered "atrophic", i.e. pathologically altered and damaged. The functional consequences of this phenomenon are poorly understood.
Results: Using computational modelling of 3D-reconstructed granule cells we show that denervation-induced dendritic atrophy also subserves homeostatic functions: By shortening their dendritic tree, granule cells compensate for the loss of inputs by a precise adjustment of excitability. As a consequence, surviving afferents are able to activate the cells, thereby allowing information to flow again through the denervated area. In addition, action potentials backpropagating from the soma to the synapses are enhanced specifically in reorganized portions of the dendritic arbor, resulting in their increased synaptic plasticity. These two observations generalize to any given dendritic tree undergoing structural changes.
Conclusions: Structural homeostatic plasticity, i.e. homeostatic dendritic remodeling, is operating in long-term denervated neurons to achieve functional homeostasis.
Abstract: Integration of synaptic currents across an extensive dendritic tree is a prerequisite for computation in the brain. Dendritic tapering away from the soma has been suggested to both equalise contributions from synapses at different locations and maximise the current transfer to the soma. To find out how this is achieved precisely, an analytical solution for the current transfer in dendrites with arbitrary taper is required. We derive here an asymptotic approximation that accurately matches results from numerical simulations. From this we then determine the diameter profile that maximises the current transfer to the soma. We find a simple quadratic form that matches diameters obtained experimentally, indicating a fundamental architectural principle of the brain that links dendritic diameters to signal transmission.
Author Summary: Neurons take a great variety of shapes that allow them to perform their different computational roles across the brain. The most distinctive visible feature of many neurons is the extensively branched network of cable-like projections that make up their dendritic tree. A neuron receives current-inducing synaptic contacts from other cells across its dendritic tree. As in the case of botanical trees, dendritic trees are strongly tapered towards their tips. This tapering has previously been shown to offer a number of advantages over a constant width, both in terms of reduced energy requirements and the robust integration of inputs at different locations. However, in order to predict the computations that neurons perform, analytical solutions for the flow of input currents tend to assume constant dendritic diameters. Here we introduce an asymptotic approximation that accurately models the current transfer in dendritic trees with arbitrary, continuously changing, diameters. When we then determine the diameter profiles that maximise current transfer towards the cell body we find diameters similar to those observed in real neurons. We conclude that the tapering in dendritic trees to optimise signal transmission is a fundamental architectural principle of the brain.
Abstract We consider the phase structure of hadronic and hadron-quark models at finite temperature and density. The basis for the hadronic part is an extension of a flavor-SU(3) ? ? ? model. We study the effect on the phase diagram by adding additional hadronic resonances to the model. With the resulting equation of state we investigate heavy-ion c... collisions using hydrodynamical simulations. In a combined approach we include quarks and the Polyakov loop field in the calculation and study chiral symmetry restoration and the deconfinement transition.
The true revolution in the age of digital neuroanatomy is the ability to extensively quantify anatomical structures and thus investigate structure-function relationships in great detail. Large-scale projects were recently launched with the aim of providing infrastructure for brain simulations. These projects will increase the need for a precise understanding of brain structure, e.g., through statistical analysis and models.
From articles in this Research Topic, we identify three main themes that clearly illustrate how new quantitative approaches are helping advance our understanding of neural structure and function. First, new approaches to reconstruct neurons and circuits from empirical data are aiding neuroanatomical mapping. Second, methods are introduced to improve understanding of the underlying principles of organization. Third, by combining existing knowledge from lower levels of organization, models can be used to make testable predictions about a higher-level organization where knowledge is absent or poor. This latter approach is useful for examining statistical properties of specific network connectivity when current experimental methods have not yet been able to fully reconstruct whole circuits of more than a few hundred neurons.
Abstract: Understanding the structure and dynamics of cortical connectivity is vital to understanding cortical function. Experimental data strongly suggest that local recurrent connectivity in the cortex is significantly non-random, exhibiting, for example, above-chance bidirectionality and an overrepresentation of certain triangular motifs. Additional evidence suggests a significant distance dependency to connectivity over a local scale of a few hundred microns, and particular patterns of synaptic turnover dynamics, including a heavy-tailed distribution of synaptic efficacies, a power law distribution of synaptic lifetimes, and a tendency for stronger synapses to be more stable over time. Understanding how many of these non-random features simultaneously arise would provide valuable insights into the development and function of the cortex. While previous work has modeled some of the individual features of local cortical wiring, there is no model that begins to comprehensively account for all of them. We present a spiking network model of a rodent Layer 5 cortical slice which, via the interactions of a few simple biologically motivated intrinsic, synaptic, and structural plasticity mechanisms, qualitatively reproduces these non-random effects when combined with simple topological constraints. Our model suggests that mechanisms of self-organization arising from a small number of plasticity rules provide a parsimonious explanation for numerous experimentally observed non-random features of recurrent cortical wiring. Interestingly, similar mechanisms have been shown to endow recurrent networks with powerful learning abilities, suggesting that these mechanism are central to understanding both structure and function of cortical synaptic wiring.
Author Summary: The problem of how the brain wires itself up has important implications for the understanding of both brain development and cognition. The microscopic structure of the circuits of the adult neocortex, often considered the seat of our highest cognitive abilities, is still poorly understood. Recent experiments have provided a first set of findings on the structural features of these circuits, but it is unknown how these features come about and how they are maintained. Here we present a neural network model that shows how these features might come about. It gives rise to numerous connectivity features, which have been observed in experiments, but never before simultaneously produced by a single model. Our model explains the development of these structural features as the result of a process of self-organization. The results imply that only a few simple mechanisms and constraints are required to produce, at least to the first approximation, various characteristic features of a typical fragment of brain microcircuitry. In the absence of any of these mechanisms, simultaneous production of all desired features fails, suggesting a minimal set of necessary mechanisms for their production.
During the last decade, Bayesian probability theory has emerged as a framework in cognitive science and neuroscience for describing perception, reasoning and learning of mammals. However, our understanding of how probabilistic computations could be organized in the brain, and how the observed connectivity structure of cortical microcircuits supports these calculations, is rudimentary at best. In this study, we investigate statistical inference and self-organized learning in a spatially extended spiking network model, that accommodates both local competitive and large-scale associative aspects of neural information processing, under a unified Bayesian account. Specifically, we show how the spiking dynamics of a recurrent network with lateral excitation and local inhibition in response to distributed spiking input, can be understood as sampling from a variational posterior distribution of a well-defined implicit probabilistic model. This interpretation further permits a rigorous analytical treatment of experience-dependent plasticity on the network level. Using machine learning theory, we derive update rules for neuron and synapse parameters which equate with Hebbian synaptic and homeostatic intrinsic plasticity rules in a neural implementation. In computer simulations, we demonstrate that the interplay of these plasticity rules leads to the emergence of probabilistic local experts that form distributed assemblies of similarly tuned cells communicating through lateral excitatory connections. The resulting sparse distributed spike code of a well-adapted network carries compressed information on salient input features combined with prior experience on correlations among them. Our theory predicts that the emergence of such efficient representations benefits from network architectures in which the range of local inhibition matches the spatial extent of pyramidal cells that share common afferent input.
Even in the absence of sensory stimulation the brain is spontaneously active. This background “noise” seems to be the dominant cause of the notoriously high trial-to-trial variability of neural recordings. Recent experimental observations have extended our knowledge of trial-to-trial variability and spontaneous activity in several directions: 1. Trial-to-trial variability systematically decreases following the onset of a sensory stimulus or the start of a motor act. 2. Spontaneous activity states in sensory cortex outline the region of evoked sensory responses. 3. Across development, spontaneous activity aligns itself with typical evoked activity patterns. 4. The spontaneous brain activity prior to the presentation of an ambiguous stimulus predicts how the stimulus will be interpreted. At present it is unclear how these observations relate to each other and how they arise in cortical circuits. Here we demonstrate that all of these phenomena can be accounted for by a deterministic self-organizing recurrent neural network model (SORN), which learns a predictive model of its sensory environment. The SORN comprises recurrently coupled populations of excitatory and inhibitory threshold units and learns via a combination of spike-timing dependent plasticity (STDP) and homeostatic plasticity mechanisms. Similar to balanced network architectures, units in the network show irregular activity and variable responses to inputs. Additionally, however, the SORN exhibits sequence learning abilities matching recent findings from visual cortex and the network's spontaneous activity reproduces the experimental findings mentioned above. Intriguingly, the network's behaviour is reminiscent of sampling-based probabilistic inference, suggesting that correlates of sampling-based inference can develop from the interaction of STDP and homeostasis in deterministic networks. We conclude that key observations on spontaneous brain activity and the variability of neural responses can be accounted for by a simple deterministic recurrent neural network which learns a predictive model of its sensory environment via a combination of generic neural plasticity mechanisms.
Tumour hypoxia plays a pivotal role in cancer therapy for most therapeutic approaches from radiotherapy to immunotherapy. The detailed and accurate knowledge of the oxygen distribution in a tumour is necessary in order to determine the right treatment strategy. Still, due to the limited spatial and temporal resolution of imaging methods as well as lacking fundamental understanding of internal oxygenation dynamics in tumours, the precise oxygen distribution map is rarely available for treatment planing. We employ an agent-based in silico tumour spheroid model in order to study the complex, localized and fast oxygen dynamics in tumour micro-regions which are induced by radiotherapy. A lattice-free, 3D, agent-based approach for cell representation is coupled with a high-resolution diffusion solver that includes a tissue density-dependent diffusion coefficient. This allows us to assess the space- and time-resolved reoxygenation response of a small subvolume of tumour tissue in response to radiotherapy. In response to irradiation the tumour nodule exhibits characteristic reoxygenation and re-depletion dynamics which we resolve with high spatio-temporal resolution. The reoxygenation follows specific timings, which should be respected in treatment in order to maximise the use of the oxygen enhancement effects. Oxygen dynamics within the tumour create windows of opportunity for the use of adjuvant chemotherapeutica and hypoxia-activated drugs. Overall, we show that by using modelling it is possible to follow the oxygenation dynamics beyond common resolution limits and predict beneficial strategies for therapy and in vitro verification. Models of cell cycle and oxygen dynamics in tumours should in the future be combined with imaging techniques, to allow for a systematic experimental study of possible improved schedules and to ultimately extend the reach of oxygenation monitoring available in clinical treatment.
Sparse coding is a popular approach to model natural images but has faced two main challenges: modelling low-level image components (such as edge-like structures and their occlusions) and modelling varying pixel intensities. Traditionally, images are modelled as a sparse linear superposition of dictionary elements, where the probabilistic view of this problem is that the coefficients follow a Laplace or Cauchy prior distribution. We propose a novel model that instead uses a spike-and-slab prior and nonlinear combination of components. With the prior, our model can easily represent exact zeros for e.g. the absence of an image component, such as an edge, and a distribution over non-zero pixel intensities. With the nonlinearity (the nonlinear max combination rule), the idea is to target occlusions; dictionary elements correspond to image components that can occlude each other. There are major consequences of the model assumptions made by both (non)linear approaches, thus the main goal of this paper is to isolate and highlight differences between them. Parameter optimization is analytically and computationally intractable in our model, thus as a main contribution we design an exact Gibbs sampler for efficient inference which we can apply to higher dimensional data using latent variable preselection. Results on natural and artificial occlusion-rich data with controlled forms of sparse structure show that our model can extract a sparse set of edge-like components that closely match the generating process, which we refer to as interpretable components. Furthermore, the sparseness of the solution closely follows the ground-truth number of components/edges in the images. The linear model did not learn such edge-like components with any level of sparsity. This suggests that our model can adaptively well-approximate and characterize the meaningful generation process.
FIAS Scientific Report 2014
(2015)
FIAS Scientific Report 2013
(2014)
FIAS Scientific Report 2012
(2013)
Intrinsic motivations drive the acquisition of knowledge and skills on the basis of novel or surprising stimuli or the pleasure to learn new skills. In so doing, they are different from extrinsic motivations that are mainly linked to drives that promote survival and reproduction. Intrinsic motivations have been implicitly exploited in several psychological experiments but, due to the lack of proper paradigms, they are rarely a direct subject of investigation. This article investigates how different intrinsic motivation mechanisms can support the learning of visual skills, such as "foveate a particular object in space", using a gaze contingency paradigm. In the experiment participants could freely foveate objects shown in a computer screen. Foveating each of two “button” pictures caused different effects: one caused the appearance of a simple image (blue rectangle) in unexpected positions, while the other evoked the appearance of an always-novel picture (objects or animals). The experiment studied how two possible intrinsic motivation mechanisms might guide learning to foveate one or the other button picture. One mechanism is based on the sudden, surprising appearance of a familiar image at unpredicted locations, and a second one is based on the content novelty of the images. The results show the comparative effectiveness of the mechanism based on image novelty, whereas they do not support the operation of the mechanism based on the surprising location of the image appearance. Interestingly, these results were also obtained with participants that, according to a post experiment questionnaire, had not understood the functions of the different buttons suggesting that novelty-based intrinsic motivation mechanisms might operate even at an unconscious level.
We investigate the properties of the QCD matter across the deconfinement phase transition. In the scope of the parton-hadron string dynamics (PHSD) transport approach, we study the strongly interacting matter in equilibrium as well as the out-of equilibrium dynamics of relativistic heavy-ion collisions. We present here in particular the results on the electromagnetic radiation, i.e. photon and dilepton production, in relativistic heavy-ion collisions and the relevant correlator in equilibrium, i.e. the electric conductivity. By comparing our calculations for the heavy-ion collisions to the available data, we determine the relative importance of the various production sources and address the possible origin of the observed strong elliptic flow ν2 of direct photons.
In the presence of a minimal length, physical objects cannot collapse to an infinite density, singular, matter point. In this paper, we consider the possible final stage of the gravitational collapse of "thick" matter layers. The energy momentum tensor we choose to model these shell-like objects is a proper modification of the source for "noncommutative geometry inspired," regular black holes. By using higher momenta of Gaussian distribution to localize matter at finite distance from the origin, we obtain new solutions of the Einstein equation which smoothly interpolates between Minkowski's geometry near the center of the shell and Schwarzschild’s spacetime far away from the matter layer. The metric is curvature singularity free. Black hole type solutions exist only for "heavy" shells; that is, M >= Me, where Me is the mass of the extremal configuration. We determine the Hawking temperature and a modified area law taking into account the extended nature of the source.
While the existence of a strongly interacting state of matter, known as “quark-gluon plasma” (QGP), has been established in heavy ion collision experiments in the past decade, the task remains to map out the transition from the hadronic matter to the QGP. This is done by measuring the dependence of key observables (such as particle suppression and elliptic flow) on the collision energy of the heavy ions. This procedure, known as "beam energy scan", has been most recently performed at the Relativistic Heavy Ion Collider (RHIC).
Utilizing a Boltzmann+hydrodynamics hybrid model, we study the collision energy dependence of initial state eccentricities and the final state elliptic and triangular flow. This approach is well suited to investigate the relative importance of hydrodynamics and hadron transport at different collision energies.
We derive the Polyakov-loop thermodynamic potential in the perturbative approach to pure SU(3) Yang-Mills theory. The potential expressed in terms of the Polyakov loop in the fundamental representation corresponds to that of the strong-coupling expansion, of which the relevant coefficients of the gluon energy distribution are specified by characters of the SU(3) group. At high temperature, the potential exhibits the correct asymptotic behavior, whereas at low temperature, it disfavors gluons as appropriate dynamical degrees of freedom. To quantify the Yang-Mills thermodynamics in confined phase, we introduce a hybrid approach which matches the effective gluon potential to that of glueballs, constrained by the QCD trace anomaly in terms of dilaton fields.
We study the impact of nonequilibrium effects on the relevant signals within a chiral fluid dynamics model including explicit propagation of the Polyakov loop. An expanding heat bath of quarks is coupled to the Langevin dynamics of the order parameter fields. The model is able to describe relaxational processes, including critical slowing down and the enhancement of soft modes near the critical point. At the first-order phase transition we observe domain formation and phase coexistence in the sigma and Polyakov loop field leading to a significant amount of clumping in the energy density. This effect gets even more pronounced if we go to systems at finite baryon density. Here the formation of high-density clusters could provide an important observable signal for upcoming experiments at FAIR and NICA.We conclude that improving our understanding of dynamical symmetry breaking is important to give realistic estimates for experimental observables connected to the QCD phase transition.
The QGP that might be created in ultrarelativistic heavy-ion collisions is expected to radiate thermal dilepton radiation. However, this thermal dilepton radiation interferes with dileptons originating from hadron decays. In the invariant mass region between the f and J=y peak (1GeV <= M l+l <=. 3GeV) the most substantial background of hadron decays originates from correlated DD¯ -meson decays. We evaluate this background using a Langevin simulation for charm quarks. As background medium we utilize the well-tested UrQMD-hybrid model. The required drag and diffusion coefficients are taken from a resonance approach. The decoupling of the charm quarks from the hot medium is performed at a temperature of 130MeV and as hadronization mechanism a coalescence approach is chosen. This model for charm quark interactions with the medium has already been successfully applied to the study of the medium modification and the elliptic flow at FAIR, RHIC and LHC energies. In this proceeding we present our results for the dilepton radiation from correlated D¯D decays at RHIC energy in comparison to PHENIX measurements in the invariant mass range between 1 and 3 GeV using different interaction scenarios. These results can be utilized to estimate the thermal QGP radiation.
As microscopic transport models usually have difficulties to deal with in-medium effects in heavy-ion collisions, we present an alternative approach that uses coarse-grained output from transport calculations with the UrQMD model to determine thermal dilepton emission rates. A four-dimensional space-time grid is set up to extract local baryon and energy densities, respectively temperature and baryon chemical potential. The lepton pair emission is then calculated for each cell of the grid using thermal equilibrium rates. In the current investigation we inlcude the medium-modified r spectral function by Eletsky et al., as well as contributions from the QGP and four-pion interactions for high collision energies. First dielectron invariant mass spectra for Au+Au collisions at 1.25 AGeV and for dimuons from In+In at 158 AGeV are shown. At 1.25 AGeV a clear enhancement of the total dilepton yield as compared to a pure transport result is observed. In the latter case, we compare our outcome with the NA60 dimuon excess data. Here a good agreement is achieved, but the yield in the low-mass tail is underestimated. In general the results show that the coarse-graining approach gives reasonable results and can cover a broad collision-energy range.
The neuroanatomical connectivity of cortical circuits is believed to follow certain rules, the exact origins of which are still poorly understood. In particular, numerous nonrandom features, such as common neighbor clustering, overrepresentation of reciprocal connectivity, and overrepresentation of certain triadic graph motifs have been experimentally observed in cortical slice data. Some of these data, particularly regarding bidirectional connectivity are seemingly contradictory, and the reasons for this are unclear. Here we present a simple static geometric network model with distance-dependent connectivity on a realistic scale that naturally gives rise to certain elements of these observed behaviors, and may provide plausible explanations for some of the conflicting findings. Specifically, investigation of the model shows that experimentally measured nonrandom effects, especially bidirectional connectivity, may depend sensitively on experimental parameters such as slice thickness and sampling area, suggesting potential explanations for the seemingly conflicting experimental results.
We study the equilibrium properties of strongly-interacting infinite parton-hadron matter, characterized by the transport coefficients such as shear and bulk viscosity and electric conductivity, and the non-equilibrium dynamics of heavy-ion collisions within the Parton-Hadron-String Dynamics (PHSD) transport approach, which incorporates explicit partonic degrees of freedom in terms of strongly interacting quasiparticles (quarks and gluons) in line with an equation of state from lattice QCD as well as the dynamical hadronization and hadronic collision dynamics in the final reaction phase. We discuss in particular the possible origin for the strong elliptic flow v2 of direct photons observed at RHIC energies.
Dendritic morphology has been shown to have a dramatic impact on neuronal function. However, population features such as the inherent variability in dendritic morphology between cells belonging to the same neuronal type are often overlooked when studying computation in neural networks. While detailed models for morphology and electrophysiology exist for many types of single neurons, the role of detailed single cell morphology in the population has not been studied quantitatively or computationally. Here we use the structural context of the neural tissue in which dendritic trees exist to drive their generation in silico. We synthesize the entire population of dentate gyrus granule cells, the most numerous cell type in the hippocampus, by growing their dendritic trees within their characteristic dendritic fields bounded by the realistic structural context of (1) the granule cell layer that contains all somata and (2) the molecular layer that contains the dendritic forest. This process enables branching statistics to be linked to larger scale neuroanatomical features. We find large differences in dendritic total length and individual path length measures as a function of location in the dentate gyrus and of somatic depth in the granule cell layer. We also predict the number of unique granule cell dendrites invading a given volume in the molecular layer. This work enables the complete population-level study of morphological properties and provides a framework to develop complex and realistic neural network models.
Currently, little is known about how synesthesia develops and which aspects of synesthesia can be acquired through a learning process. We review the increasing evidence for the role of semantic representations in the induction of synesthesia, and argue for the thesis that synesthetic abilities are developed and modified by semantic mechanisms. That is, in certain people semantic mechanisms associate concepts with perception-like experiences—and this association occurs in an extraordinary way. This phenomenon can be referred to as “higher” synesthesia or ideasthesia. The present analysis suggests that synesthesia develops during childhood and is being enriched further throughout the synesthetes’ lifetime; for example, the already existing concurrents may be adopted by novel inducers or new concurrents may be formed. For a deeper understanding of the origin and nature of synesthesia we propose to focus future research on two aspects: (i) the similarities between synesthesia and ordinary phenomenal experiences based on concepts; and (ii) the tight entanglement of perception, cognition and the conceptualization of the world. Importantly, an explanation of how biological systems get to generate experiences, synesthetic or not, may have to involve an explanation of how semantic networks are formed in general and what their role is in the ability to be aware of the surrounding world.
Information theory allows us to investigate information processing in neural systems in terms of information transfer, storage and modification. Especially the measure of information transfer, transfer entropy, has seen a dramatic surge of interest in neuroscience. Estimating transfer entropy from two processes requires the observation of multiple realizations of these processes to estimate associated probability density functions. To obtain these necessary observations, available estimators typically assume stationarity of processes to allow pooling of observations over time. This assumption however, is a major obstacle to the application of these estimators in neuroscience as observed processes are often non-stationary. As a solution, Gomez-Herrero and colleagues theoretically showed that the stationarity assumption may be avoided by estimating transfer entropy from an ensemble of realizations. Such an ensemble of realizations is often readily available in neuroscience experiments in the form of experimental trials. Thus, in this work we combine the ensemble method with a recently proposed transfer entropy estimator to make transfer entropy estimation applicable to non-stationary time series. We present an efficient implementation of the approach that is suitable for the increased computational demand of the ensemble method's practical application. In particular, we use a massively parallel implementation for a graphics processing unit to handle the computationally most heavy aspects of the ensemble method for transfer entropy estimation. We test the performance and robustness of our implementation on data from numerical simulations of stochastic processes. We also demonstrate the applicability of the ensemble method to magnetoencephalographic data. While we mainly evaluate the proposed method for neuroscience data, we expect it to be applicable in a variety of fields that are concerned with the analysis of information transfer in complex biological, social, and artificial systems.
Poster presentation at The Twenty Third Annual Computational Neuroscience Meeting: CNS*2014 Québec City, Canada. 26-31 July 2014: We study random strongly heterogeneous recurrent networks of firing rate neurons, introducing the notion of cohorts: groups of co-active neurons, who compete for firing with one another and whose presence depends sensitively on the structure of the input. The identities of neurons recruited to and dropped from an active cohort changes smoothly with varying input features. We search for network parameter regimes in which the activation of cohorts is robust yet easily switchable by the external input and which exhibit large repertoires of different cohorts. We apply these networks to model the emergence of orientation and direction selectivity in visual cortex. We feed these random networks with a set of harmonic inputs that vary across neurons only in their temporal phase, mimicking the feedforward drive due to a moving grating stimulus. The relationship between the phases that carries the information about the orientation of the stimulus determines which cohort of neurons is activated. As a result the individual neurons acquire non-monotonic orientation tuning curves which are characterized by high orientation and direction selectivity. This mechanism of emergence for direction selectivity differs from the classical motion detector scheme, which is based on the nonlinear summation of the time-shifted inputs. In our model these two mechanisms coexist in the same network, but can be distinguished by their different frequency and contrast dependences. In general, the mechanism we are studying here converts temporal phase sequence into population activity and could therefore be used to extract and represent also various other relevant stimulus features.
The YaeJ protein is a codon-independent release factor with peptidyl-tRNA hydrolysis (PTH) activity, and functions as a stalled-ribosome rescue factor in Escherichia coli. To identify residues required for YaeJ function, we performed mutational analysis for in vitro PTH activity towards rescue of ribosomes stalled on a non-stop mRNA, and for ribosome-binding efficiency. We focused on residues conserved among bacterial YaeJ proteins. Additionally, we determined the solution structure of the GGQ domain of YaeJ from E. coli using nuclear magnetic resonance spectroscopy. YaeJ and a human homolog, ICT1, had similar levels of PTH activity, despite various differences in sequence and structure. While no YaeJ-specific residues important for PTH activity occur in the structured GGQ domain, Arg118, Leu119, Lys122, Lys129 and Arg132 in the following C-terminal extension were required for PTH activity. All of these residues are completely conserved among bacteria. The equivalent residues were also found in the C-terminal extension of ICT1, allowing an appropriate sequence alignment between YaeJ and ICT1 proteins from various species. Single amino acid substitutions for each of these residues significantly decreased ribosome-binding efficiency. These biochemical findings provide clues to understanding how YaeJ enters the A-site of stalled ribosomes.
Background: Simple peak-picking algorithms, such as those based on lineshape fitting, perform well when peaks are completely resolved in multidimensional NMR spectra, but often produce wrong intensities and frequencies for overlapping peak clusters. For example, NOESY-type spectra have considerable overlaps leading to significant peak-picking intensity errors, which can result in erroneous structural restraints. Precise frequencies are critical for unambiguous resonance assignments.
Results: To alleviate this problem, a more sophisticated peaks decomposition algorithm, based on non-negative matrix factorization (NMF), was developed. We produce peak shapes from Fourier-transformed NMR spectra. Apart from its main goal of deriving components from spectra and producing peak lists automatically, the NMF approach can also be applied if the positions of some peaks are known a priori, e.g. from consistently referenced spectral dimensions of other experiments.
Conclusions: Application of the NMF algorithm to a three-dimensional peak list of the 23 kDa bi-domain section of the RcsD protein (RcsD-ABL-HPt, residues 688-890) as well as to synthetic HSQC data shows that peaks can be picked accurately also in spectral regions with strong overlap.
Self-organization is thought to play an important role in structuring nervous systems. It frequently arises as a consequence of plasticity mechanisms in neural networks: connectivity determines network dynamics which in turn feed back on network structure through various forms of plasticity. Recently, self-organizing recurrent neural network models (SORNs) have been shown to learn non-trivial structure in their inputs and to reproduce the experimentally observed statistics and fluctuations of synaptic connection strengths in cortex and hippocampus. However, the dynamics in these networks and how they change with network evolution are still poorly understood. Here we investigate the degree of chaos in SORNs by studying how the networks' self-organization changes their response to small perturbations. We study the effect of perturbations to the excitatory-to-excitatory weight matrix on connection strengths and on unit activities. We find that the network dynamics, characterized by an estimate of the maximum Lyapunov exponent, becomes less chaotic during its self-organization, developing into a regime where only few perturbations become amplified. We also find that due to the mixing of discrete and (quasi-)continuous variables in SORNs, small perturbations to the synaptic weights may become amplified only after a substantial delay, a phenomenon we propose to call deferred chaos.
Evidence from anatomical and functional imaging studies have highlighted major modifications of cortical circuits during adolescence. These include reductions of gray matter (GM), increases in the myelination of cortico-cortical connections and changes in the architecture of large-scale cortical networks. It is currently unclear, however, how the ongoing developmental processes impact upon the folding of the cerebral cortex and how changes in gyrification relate to maturation of GM/WM-volume, thickness and surface area. In the current study, we acquired high-resolution (3 Tesla) magnetic resonance imaging (MRI) data from 79 healthy subjects (34 males and 45 females) between the ages of 12 and 23 years and performed whole brain analysis of cortical folding patterns with the gyrification index (GI). In addition to GI-values, we obtained estimates of cortical thickness, surface area, GM and white matter (WM) volume which permitted correlations with changes in gyrification. Our data show pronounced and widespread reductions in GI-values during adolescence in several cortical regions which include precentral, temporal and frontal areas. Decreases in gyrification overlap only partially with changes in the thickness, volume and surface of GM and were characterized overall by a linear developmental trajectory. Our data suggest that the observed reductions in GI-values represent an additional, important modification of the cerebral cortex during late brain maturation which may be related to cognitive development.
The efficient coding hypothesis posits that sensory systems of animals strive to encode sensory signals efficiently by taking into account the redundancies in them. This principle has been very successful in explaining response properties of visual sensory neurons as adaptations to the statistics of natural images. Recently, we have begun to extend the efficient coding hypothesis to active perception through a form of intrinsically motivated learning: a sensory model learns an efficient code for the sensory signals while a reinforcement learner generates movements of the sense organs to improve the encoding of the signals. To this end, it receives an intrinsically generated reinforcement signal indicating how well the sensory model encodes the data. This approach has been tested in the context of binocular vison, leading to the autonomous development of disparity tuning and vergence control. Here we systematically investigate the robustness of the new approach in the context of a binocular vision system implemented on a robot. Robustness is an important aspect that reflects the ability of the system to deal with unmodeled disturbances or events, such as insults to the system that displace the stereo cameras. To demonstrate the robustness of our method and its ability to self-calibrate, we introduce various perturbations and test if and how the system recovers from them. We find that (1) the system can fully recover from a perturbation that can be compensated through the system's motor degrees of freedom, (2) performance degrades gracefully if the system cannot use its motor degrees of freedom to compensate for the perturbation, and (3) recovery from a perturbation is improved if both the sensory encoding and the behavior policy can adapt to the perturbation. Overall, this work demonstrates that our intrinsically motivated learning approach for efficient coding in active perception gives rise to a self-calibrating perceptual system of high robustness.
Tumour cells show a varying susceptibility to radiation damage as a function of the current cell cycle phase. While this sensitivity is averaged out in an unperturbed tumour due to unsynchronised cell cycle progression, external stimuli such as radiation or drug doses can induce a resynchronisation of the cell cycle and consequently induce a collective development of radiosensitivity in tumours. Although this effect has been regularly described in experiments it is currently not exploited in clinical practice and thus a large potential for optimisation is missed. We present an agent-based model for three-dimensional tumour spheroid growth which has been combined with an irradiation damage and kinetics model. We predict the dynamic response of the overall tumour radiosensitivity to delivered radiation doses and describe corresponding time windows of increased or decreased radiation sensitivity. The degree of cell cycle resynchronisation in response to radiation delivery was identified as a main determinant of the transient periods of low and high radiosensitivity enhancement. A range of selected clinical fractionation schemes is examined and new triggered schedules are tested which aim to maximise the effect of the radiation-induced sensitivity enhancement. We find that the cell cycle resynchronisation can yield a strong increase in therapy effectiveness, if employed correctly. While the individual timing of sensitive periods will depend on the exact cell and radiation types, enhancement is a universal effect which is present in every tumour and accordingly should be the target of experimental investigation. Experimental observables which can be assessed non-invasively and with high spatio-temporal resolution have to be connected to the radiosensitivity enhancement in order to allow for a possible tumour-specific design of highly efficient treatment schedules based on induced cell cycle synchronisation.
Author Summary: The sensitivity of a cell to a dose of radiation is largely affected by its current position within the cell cycle. While under normal circumstances progression through the cell cycle will be asynchronous in a tumour mass, external influences such as chemo- or radiotherapy can induce a synchronisation. Such a common progression of the inner clock of the cancer cells results in the critical dependence on the effectiveness of any drug or radiation dose on a suitable timing for its administration. We analyse the exact evolution of the radiosensitivity of a sample tumour spheroid in a computer model, which enables us to predict time windows of decreased or increased radiosensitivity. Fractionated radiotherapy schedules can be tailored in order to avoid periods of high resistance and exploit the induced radiosensitivity for an increase in therapy efficiency. We show that the cell cycle effects can drastically alter the outcome of fractionated irradiation schedules in a spheroid cell system. By using the correct observables and continuous monitoring, the cell cycle sensitivity effects have the potential to be integrated into treatment planing of the future and thus to be employed for a better outcome in clinical cancer therapies.