SemOpenAlex |

SemOpenAlex

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385698947> ?p ?o ?g. }

Showing items 1 to 59 of 59 with 100 items per page.

W4385698947 abstract "Full text Figures and data Side by side Abstract Editor's evaluation Introduction Results Discussion Methods Data availability References Decision letter Author response Article and author information Metrics Abstract Oscillations occurring simultaneously in a given area represent a physiological unit of brain states. They allow for temporal segmentation of spikes and support distinct behaviors. To establish how multiple oscillatory components co-vary simultaneously and influence neuronal firing during sleep and wakefulness in mice, we describe a multivariate analytical framework for constructing the state space of hippocampal oscillations. Examining the co-occurrence patterns of oscillations on the state space, across species, uncovered the presence of network constraints and distinct set of cross-frequency interactions during wakefulness compared to sleep. We demonstrated how the state space can be used as a canvas to map the neural firing and found that distinct neurons during navigation were tuned to different sets of simultaneously occurring oscillations during sleep. This multivariate analytical framework provides a window to move beyond classical bivariate pipelines for investigating oscillations and neuronal firing, thereby allowing to factor-in the complexity of oscillation–population interactions. Editor's evaluation Traditional approaches for the study of brain oscillations are typically based on analyzing spectral features of individual oscillations (univariate methods) or the power and phase relationship between two oscillations (bivariate methods). This manuscript presents a different, multivariate, approach to simultaneously analyze interactions between multiple oscillations and applied it to rodent hippocampal LFPs. This innovative and important method provides a comprehensive and convincing approach to characterize oscillatory network states, opening new avenues for studying the complex interactions that characterize neural circuit dynamics. https://doi.org/10.7554/eLife.80263.sa0 Decision letter Reviews on Sciety eLife's review process Introduction Oscillations, generated by large neuronal ensembles, are a hallmark of the mammalian brain. They are well preserved during evolution (Buzsáki and Draguhn, 2004) and have been suggested to play a key role in high cognitive functions (Uhlhaas and Singer, 2010). They enable the synchronization of neural activity within and across brain regions, thus promoting the precise temporal coordination of neural processes underlying memory, perception, and behavior (Colgin and Moser, 2010). Their disruption leads to cognitive and sensorimotor deficits associated with several neuropsychiatric diseases (Uhlhaas et al., 2011). Oscillations occurring at different frequency bands (from 0.5 to 200 Hz), interact with each other through hierarchical cross-frequency coupling (González et al., 2020; Jensen and Colgin, 2007; Tort et al., 2008). Such interactions are linked to different computational operations or different phases in a computational operation and are thought to characterize distinct brain states supporting various aspects of behavior (McCormick et al., 2020). In the hippocampus, theta cycles are often nested with bouts of faster oscillations in the gamma frequency and are thought to be related to different stages of memory processing (Lopes-Dos-Santos et al., 2018). Specifically, during wakefulness, the CA1 region is characterized by cross-frequency coupling between theta and distinct gamma oscillations, which exhibit their peak power at distinct phases of ongoing slow oscillations (Belluscio et al., 2012; Buzsáki and Wang, 2012; Colgin et al., 2009; Csicsvari et al., 2003; Scheffer-Teixeira et al., 2012; Yamamoto et al., 2014). The slow, medium, and fast gamma oscillations, emerging from stratum radiatum (SR), stratum lacunosum moleculare (SLM), and stratum pyramidale (SP), respectively, are linked to network activity in entorhinal cortex, CA3 and CA1 region, respectively (Colgin et al., 2009; Fernández-Ruiz et al., 2017; Guardamagna et al., 2023; Lasztóczi and Klausberger, 2016; Schomburg et al., 2014). During sleep, theta and medium gamma oscillations dominate the rapid eye movement REM sleep, whereas delta, beta frequency band, and ripples or fast gamma (100–200 Hz), linked to memory consolidation processes, dominate the non-REM sleep (Battaglia et al., 2011; Buzsáki, 1989; Genzel et al., 2014; Roumis and Frank, 2015). The composition of hippocampal states by distinct set of slow and fast oscillations reflects synchrony among different brain areas and corresponds to distinct functional states of the hippocampal network (Colgin, 2015). However, despite extensive research on simultaneously occurring oscillations, very little is known about the dynamic variations in the composition of hippocampal network state with time and behavior. Oscillations are generated by neuronal population, and they in turn are known to modulate the firing activity of individual cells (Benchenane et al., 2010; Hulse et al., 2017; O’Keefe and Recce, 1993). Classically, neuronal firing has been characterized in the context of a single-frequency band during distinct brain states (Fox et al., 1986; Henze et al., 2000; Hafting et al., 2008; van Wingerden et al., 2010; Siapas et al., 2005). However, the classical analytical methods, due to their bivariate nature, are limited to examining the firing activity of neurons in relation to power or phase or frequency of a specific oscillation only, thus neglecting how multiple oscillations, occurring simultaneously in a given region, modulate the neuronal firing in a combinatorial manner. Here, a novel analytical approach has been described to investigate how simultaneously occurring network oscillations dynamically contribute to the composition of hippocampal states and how they influence the neuronal firing. To this aim, multisite local field potentials (LFPs), recorded from the CA1 region of the dorsal hippocampus of freely moving mice, were used to construct the network state space during wakefulness and sleep. This method has allowed to create a compact representation of the state of multiple oscillatory processes, distinct in frequency and anatomical localization. We used this compact representation for studying various oscillations, their intrinsic organization, their temporal progression, and their simultaneous influence on distinct neuronal ensembles and behavior. In addition, the state space provided a window for observing the hippocampal population in the context of the network in which they are embedded. This allowed us to examine how cells fire as a function of the network state and how multiple oscillations simultaneously modulate cell firing. As a proof of concept, this approach has been applied to datasets recorded, in similar experimental conditions, from another animal species (rats), in Dr. Gyorgy Buzsaki’s lab at NYU (hc-11 dataset, crncs.org; Grosmark et al., 2016; Grosmark and Buzsáki, 2016). Lastly, the network state space framework has been applied to study alterations in the organization of hippocampal oscillations in mice lacking neuroligin 3 (NLG3 KO; neuroligin-3 knock out), an animal model of autism (Baudouin et al., 2012). Results Experimental paradigm and construction of the network state space On a single day, we recorded brain activity in freely moving mice (n = 4) during four consecutive trials: (1) Sleep1, (2) Novel Arena-1 Exploration, (3) Novel Arena-2 Exploration, and (4) Sleep2, (Figure 1A). Layer-resolved LFP were extracted from the CA1 region of the dorsal hippocampus using multisite silicon probes (Figure 1B; Guardamagna et al., 2022). LFPs from SP, SR, and SLM layers were used to extract signals in the following frequency bands: delta (1–5 Hz), theta (6–10 Hz), beta (10–20 Hz), slow gamma (20–45 Hz), medium gamma (60–90 Hz), and fast gamma (100–200 Hz). Each of these 18 frequency bands (six from each layer, Figure 1C) were used to compute median power in nonoverlapping bins of 200 milliseconds and then smoothed using Gaussian kernel. The resulting 18 power time series across all four trials combined (divided into, say, N bins of 200 ms each) form a cloud of N points in 18D space, which we refer to as the network state space (Figure 2A). Each point in the cloud represents the state of the network (i.e., the power configuration of the 18 frequency bands from three layers) at a given time. Uniform Manifold Approximation and Projection (UMAP, McInnes et al., 2020) was employed to reduce the dimensionality of the network state space from 18D to 2D (Figures 1C and 2A). A similar state space was constructed using the power in current source density (CSD) signals instead of LFPs (Figure 1—figure supplement 1). We next characterized some fundamental properties of the network state space during sleep and wakefulness. Figure 1 with 2 supplements see all Download asset Open asset Experimental paradigm and pipeline for construction of the network state space. (A) Representative four-trial sequence for recording during sleep and awake exploration (top to bottom). (B) Representative traces of local field potential (LFP) recorded from various layers of dorsal CA1 using silicon probe. (C) Pipeline for the construction of network state space. See also Figure 1—figure supplements 1 and 2. Figure 2 with 4 supplements see all Download asset Open asset Restricted occupancy of the network state space during wakefulness. (A) Network state space for all four trials combined. (B) Trial-specific states (colored) and unvisited states (gray) on the network state space. (C) Fraction of state space occupied during sleep 0.81±0.01 and awake 0.26±0.01 trials, suggesting significant restrictions during awake trials (n = 8 sleep and awake trials from four mice. p=0.0002, Mann–Whitney test). (D) Distribution of difference in fraction of state space occupied during sleep and awake trials (Sleep-Awake) of surrogate data and observed data. See also Figure 2—figure supplements 1–4. Restricted occupancy of the network state space during wakefulness We first examined the occupancy of the network state space during sleep and awake exploration by overlaying trial specific states (states that network visits in a given trial) on the network state space (Figure 2A and B). We observed that during awake trials the activity of the network was restricted to a subset of the states, whereas during sleep periods, it occupied a significantly larger area of the state space. This was quantified by computing the fraction of the state space occupied during each trial. In all four mice (eight sleep and eight awake trials), we observed a significant restriction (Figure 2C, Figure 2—figure supplement 1) of the state space’s occupancy during wakefulness compared to sleep (0.26±0.01versus0.81±0.01) . The observed restriction on the state space was further compared against a surrogate data generated by randomly shuffling the binned power in all frequency bands (see ‘Methods’). We found that the observed difference in occupancy between sleep and awake trials was significantly different from those computed in surrogate data (Figure 2D). To assess whether the observed restricted occupancy was due to smaller trial duration of awake trails (20 min) compared to sleep (60 min), we performed control analysis by computing occupancy on the state space obtained by randomly selecting equal number of network states from awake and sleep trials. We observed similar restrictions (Figure 2—figure supplement 1C), suggesting that the restricted occupancy during wakefulness is independent of trial duration. Next, we assessed whether the observed restrictions on the state space are driven by large fluctuations in low-frequency bands, which may weigh preponderantly in the UMAP. We computed the state space with normalized features (Figure 2—figure supplement 2) as well as obtained the state space projection using an alternative dimensionality reduction approach (principal component analysis; Figure 2—figure supplement 4). Similar patterns of restrictions on the network state space, during wakefulness, across all different projections were observed. To further validate our findings, we computed the network state space and occupancy for the datasets recorded from freely moving rats in similar experimental conditions (Figure 2—figure supplement 3) and observed remarkable similarities in restrictions on the state space, suggesting that functional organization of network oscillations during sleep and wakefulness is conserved across species. In addition, we observed that the network occupies the same set of states when animals explore distinct arenas (Figure 2B, Figure 2—figure supplement 1). Hence, the restrictions are primarily dependent on the behavioral state of the animal rather than environmental factors. Characterization of restrictions on the network state space during wakefulness To characterize the network restrictions and visualize how different hippocampal rhythms vary simultaneously during sleep and awake trials, we overlaid the power of each oscillation on the network state space (Figure 3A). While classical analytical approaches employ bivariate methods by evaluating each oscillation individually across sleep and awake trials (Figure 3B), the network state space employs multivariate approach and highlights how fluctuations in power for one oscillation link to the general oscillatory state of the network. This allows visualizing how the power of one oscillation varies in the context of all other oscillations. In addition, it underscores the regime of operation for different oscillations during sleep and wakefulness. For instance, the restricted subspace of the state space occupied during wakefulness corresponds to states with low delta power, moderate beta power, and moderate to high gamma power. However, during sleep, network oscillations exhibit broader range of operations since the power of each oscillation varies from its minima to its maxima as evident in the overlay maps (Figure 3A). This characteristic distribution during sleep was further used to determine the localization of REM, non-REM, and intermediate sleep states on the network state space. Figure 3 Download asset Open asset Characterization of restrictions on the network state space during wakefulness. (A) Distribution of power on state space for six oscillations of layer stratum pyramidale: delta (1–5 Hz), theta (6–10 Hz), beta (10–20 Hz), slow gamma (20–45 Hz), medium gamma (60–90 Hz), and fast gamma (100–200 Hz). Each oscillation is individually color-scaled to its respective minimum and maximum power. Unvisited states are in gray. The overlay maps demonstrate how the power of each oscillation varies on the state space and across sleep and awake trials. (B) Mean power comparison between awake and sleep trials using classical approach (n = 4 mice). All p<0.05, except for theta (p=0.11), Mann–Whitney test. Distinct localization of REM and non-REM sleep states on state space Sleep trials were further classified into REM, non-REM, and intermediate states using classical definitions (see ‘Methods’). We overlaid the theta/delta power ratio for each state on the network state space (Figure 4A). States with high theta/delta ratio were identified as REM (Figure 4A, C and D, Figure 4—figure supplement 1). Similarly, non-REM sleep states were detected by overlaying delta × beta power product for each state on the state space (Figure 4B–D). The remaining states were classified as intermediate sleep states. These overlay maps allowed to visualize how functionally distinct sleep states are localized on the state space and they were further used to study how network properties varies in distinct regions of the state space. Visual inspection of power overlay maps (Figure 3A and REM and non-REM states localization maps in Figure 4A and B) revealed that REM states are characterized by moderate to high power in beta, slow, medium, and fast gamma oscillatory bands simultaneously with theta oscillations, whereas non-REM states are characterized by the presence of moderate to high power in theta, slow, and fast gamma oscillations along with delta and beta. Notably, the medium gamma oscillations operate in low-power mode during non-REM sleep. This co-occurrence of network oscillations is formally quantified in later sections. Figure 4 with 1 supplement see all Download asset Open asset Localization of rapid eye movement (REM) and non-REM sleep states and density on the network state space. (A) Left: standardized theta/delta ratio from pyramidal layer overlaid on state space. Right: identified REM states (magenta colored). (B) Left: standardized delta × beta product from pyramidal layer overlaid on state space. Right: identified non-REM states (cyan colored). (C) Representative REM and non-REM states on the network state space. (D) Corresponding REM and non-REM local field potential (LFP) from pyramidal layer. (E) Density overlaid on state space across all four trials (unvisited states are colored gray). (F) Representative change in density map across two sleep trials (Sleep2 – Sleep1). (G) Left: comparison of median REM density between two sleep trials (n = 4 mice); (Sleep1: 0.0076±5e-4 v/s Sleep2: 0.0075 ± 0.001, p=0.88 Mann–Whitney test); Center: comparison of median non-REM density between two sleep trials (n = 4 mice); (Sleep1: 0.039 ± 0.001 v/s Sleep2: 0.048 ± 0.002, p<0.05, Mann–Whitney test); right: median density comparison between and sleep and awake trials (n = 8 sleep and awake trials, four mice) (0.03 ± 0.0005 v/s 0.13 ± 0.009, p=0.0002 Mann–Whitney test). See also Figure 4—figure supplement 1. Increased density of non-REM states associated with theta and gamma oscillations after exploration We next computed network density that represents the fraction of time the network spends in each bin on the network state space (Figure 4E). It is measured in the units: number of state visits/bin/second. Density on the state space computed during awake trials measures the normalized occurrence of corresponding behaviors during exploration whereas density computed during sleep trials measures the time spent by the network in various sleep states (REM, non-REM). Mirroring the smaller occupied state space region, the median density of awake trials was significantly higher than that detected during sleep (0.13 ± 0.009 versus 0.03 ± 0.0005). We next investigated the density across sleep trials (Sleep1 and Sleep2) by comparing the median density of REM and non-REM states between the two sleep trials (Figure 4F). We observed an increase in the median density of non-REM states during post-exploration sleep (Figure 4G: center; Sleep1: 0.039 ± 0.001versus Sleep2: 0.048 ± 0.002) while the density of REM sleep remained identical (Figure 4C, left; Sleep1: 0.0076±5e-4 versus Sleep2: 0.0075 ± 0.001). In particular, non-REM states in Sleep2 tended to concentrate in a region of increased power in the delta and beta bands, which could be the results of increased interactions with cortical activity modulated in the same range. It is also likely that such effect was induced by the exposure to relevant behavioral experience. In fact, changes in density of individual oscillations after learning have been reported using traditional analytical methods and are thought to support memory consolidation (Bakker et al., 2015; Eschenko et al., 2008; Eschenko et al., 2006). Nevertheless, while traditional methods provide information about individual components, the novel approach used here provides additional information about the combinatorial shift in the dynamics of network oscillations after learning or exploration. Thus, it provides the basis for identifying how coordinated activity among different oscillations supports memory consolidation processes as those occurring during non-REM sleep after exploration, which cannot be elucidated using traditional analytical methods. In this study, although REM and non-REM sleep states were identified using classical methods, the state space allowed quantifying changes in density of all other oscillations such as theta, slow, medium, and fast gamma that occur simultaneously with delta and beta bands (from power distribution maps [Figure 3A] and REM and non-REM maps [Figure 4A]). Thus, identification of sleep states on state space provides additional information about oscillation-specific changes across sleep trials. However, whether other oscillations may be statistically related to delta and beta during non-REM states of sleep and whether such correlation depends on the state of the network remain to be elucidated. State and layer-dependent coupling of hippocampal oscillations To investigate state-dependent coupling among various oscillatory processes, we computed pairwise correlation matrix using binned power in all 18 frequency bands (see ‘Methods’). Each correlation matrix represents how various bands are coupled during sleep and awake trials (Figure 5A). We observed distinct configuration of frequency band’s correlations during sleep and awake trails (Figure 5B). We further investigated how coupling of bands varies on the state space by binning the state space (Figure 5C) and computed correlation matrix in each bin. We classified the bins into one of the following states: awake, REM, non-REM, and intermediate sleep (see ‘Methods’). We employed UMAP to visualize the variability across correlation matrices. (Figure 5D) and observed distinct coupling configurations for awake, REM, and non-REM states (Figure 5E). During REM, we found increased coupling (positive correlation) among beta and slow gamma oscillations as well as among medium and fast gamma bands accompanied by decoupling (negative correlation) with delta. This organization among oscillations was altered during non-REM sleep as all oscillations except delta were coupled. Gamma segregation and delta decoupling offer a picture of hippocampal REM sleep as being more akin to awake locomotion (with the major difference of a stronger medium gamma presence) while also suggesting a substantial independence from cortical slow oscillations. On the other hand, the across-scale coherence of non-REM sleep is consistent with this sleep stage being dominated by brain-wide collective fluctuations engaging oscillations at every range. Distinct cross-frequency coupling among various individual pairs of oscillations such as theta-gamma, delta-gamma, etc., have been already reported (Bandarabadi et al., 2019; Clemens et al., 2009; Hammer et al., 2021; Scheffzük et al., 2011). However, computing cross-frequency coupling on the state space provides the additional information on how multiple oscillations, obtained from distinct CA1 hippocampal layers (stratum pyramidale, stratum radiatum, and stratum lacunosum moleculare), are coupled with each other during distinct states of sleep and wakefulness. Furthermore, projecting the correlation matrices on 2D plane provides a compact tool that allows to visualize the cross-frequency interactions among various hippocampal oscillations. Altogether, this approach reveals the complex nature of coupling dynamics occurring in hippocampus during distinct behavioral states. Figure 5 Download asset Open asset State and layer-dependent coupling of hippocampal oscillations. (A) Representative pairwise correlation among 18 oscillations (D, delta; T, theta; B, beta; SG, slow gamma; MG, medium gamma; FG, fast gamma) from three layers (stratum pyramidale [SP], stratum radiatum [SR], stratum lacunosum moleculare [SLM]). Rows are arranged by combining all three layers for each oscillation. (B) Correlation matrix space generated by using matrices in (A) from all the mice (total 16 points, 8 sleep and 8 awake trials from four mice). Each point represents the correlation matrix of oscillations. The separation of sleep and awake points in the correlation matrix space suggests distinct nature of coupling among oscillations during sleep and wakefulness. (C) Binned state space (gray) for a sleep trial. Representative rapid eye movement (REM) and non-REM bins highlighted (black) and their corresponding correlation matrix of oscillations. (D) Correlation matrix space constructed using matrices from all bins collected from all trials and all animals. Each point on correlation matrix space projection represents a correlation matrix. (E) Bin status (awake, REM, non-REM, intermediate sleep) overlaid on correlation matrix space, exhibiting state dependent coupling of hippocampal oscillations (p<0.0001 using multivariate ANOVA). After characterizing the static properties of state space such as occupancy, power distribution, localization of sleep states, network state density, and coupling among oscillations, in the following analyses we addressed the dynamic properties of state space such as network flow, state transitions, and speed on state space. Alterations in sleep state transitions after exploration/learning To investigate how learning during exploration affects the consequent sleep, we characterized state transition patterns during Sleep1 and Sleep2. By plotting outgoing trajectories on the network state space (Figure 6A, Figure 6—figure supplement 1), we visualized state transitions during sleep. We quantified the probability of state transitions among various sleep states (REM, non-REM, etc.) in a transition matrix (Figure 6B). We calculated changes in transition probabilities across sleep trials by measuring average absolute change in probability (AACP, see ‘Methods’). This allowed to quantify the amount of sleep state transitions altered after exploration/learning. To assess whether these changes in transition probabilities of sleep states are random or not, we generated 1000 pairs of sleep and awake trials by randomly shuffling state transitions from sleep data and computed transition matrices, difference matrix, and their corresponding AACP. The AACP value of real data was then compared to the distribution of randomly shuffled trials. We observed that the AACP value of real data was beyond the 95th percentile of the distribution of shuffled data (Figure 6D), Altogether, this data highlights the specific alterations in general sleep architecture and hippocampal oscillatory landscape following learning/novel exploration. Figure 6 with 2 supplements see all Download asset Open asset Alterations in sleep state transitions after exploration. (A) Incoming and outgoing trajectories for two representative bins on state space during a sleep trial. Each trajectory spans 1 s in time before (incoming) and after (outgoing) the occurrence of a given state in each bin. (B) State transition matrices for two sleep trials: pre-exploration sleep (Sleep1) and post-exploration sleep (Sleep2). (C) Change in transition probabilities across two sleep trials obtained by subtracting two transition matrices (Sleep2 – Sleep1). (D) Comparison of observed sleep data’s average absolute change in probability with shuffled sleep data. Average absolute change in probability (AACP) across sleep trials for observed data = 0.021, 95th percentile of shuffled data = 0.007. (E) AACP across sleep trials for transitions originating from rapid eye movement (REM), non-REM, and intermediate (INT) state to REM, non-REM, and intermediate sleep state (REM: p<0.0001; non-REM: p<0.0001; INT: p=0.38, one-way ANOVA). (F) Schematic diagram of absolute change in transition probabilities across sleep trials. Arrow’s thickness corresponds to absolute change in transition probabilities. See also Figure 6—figure supplements 1 and 2. Intra-state sleep transitions are more plastic than inter-state transitions We further examined which transitions on the state space are significantly altered across sleep trials. We computed AACP specifically for transition from REM/non-REM/intermediate sleep state to REM/non-REM/intermediate state. We found that transitions occurring from REM-to-REM sleep and non-REM-to-non-REM sleep (intra-state transitions) are more vulnerable to plasticity after exploration as compared to inter-state transitions (such as non-REM to REM, REM-to-intermediate, etc.) (Figure 6E and F). These changes in intra-state transitions were observed to be beyond randomness (Figure 6—figure supplement 2A and B), indicating a specificity in plastic changes in state transitions after exploration. In particular, while the average REM period duration is unaltered after exploration (Figure 4G), REM temporal structure is reorganized. In fact, increased probability of REM to REM transitions indicates a significant prolongation of REM bout duration. Similarly, the increase in non-REM to non-REM transition probability reflects an increased duration of non-REM bouts. Therefore, environment exploration was accompanied by an increased separation between REM and non-REM periods, possibly as a response to increased computational demands. More in general, the network state space allows to characterize the state transitions in hippocampus and how they are affected by novel experience or learning. By observing the state transition patterns, this analytical framework allows to detect and identify state-specific changes in the hippocampal oscillatory dynamics, beyond the possibilities offered by more traditional univariate and bivariate methods. We next investigated how fast the network flows on the state space and assessed whether the speed is uniform or whether it exhibits specific region-dependent characteristics. Stabilization on state space during transition to REM Coverage speed determines how fast the network sweeps the area of the state space. To calculate the coverage speed" @default.
W4385698947 created "2023-08-10" @default.
W4385698947 creator A5009988687 @default.
W4385698947 creator A5054923894 @default.
W4385698947 creator A5063892428 @default.
W4385698947 creator A5084451044 @default.
W4385698947 creator A5086494201 @default.
W4385698947 creator A5087367544 @default.
W4385698947 date "2023-06-09" @default.
W4385698947 modified "2023-09-23" @default.
W4385698947 title "Author response: State-dependent coupling of hippocampal oscillations" @default.
W4385698947 doi "https://doi.org/10.7554/elife.80263.sa2" @default.
W4385698947 hasPublicationYear "2023" @default.
W4385698947 type Work @default.
W4385698947 citedByCount "0" @default.
W4385698947 crossrefType "peer-review" @default.
W4385698947 hasAuthorship W4385698947A5009988687 @default.
W4385698947 hasAuthorship W4385698947A5054923894 @default.
W4385698947 hasAuthorship W4385698947A5063892428 @default.
W4385698947 hasAuthorship W4385698947A5084451044 @default.
W4385698947 hasAuthorship W4385698947A5086494201 @default.
W4385698947 hasAuthorship W4385698947A5087367544 @default.
W4385698947 hasBestOaLocation W43856989471 @default.
W4385698947 hasConcept C11413529 @default.
W4385698947 hasConcept C121332964 @default.
W4385698947 hasConcept C127413603 @default.
W4385698947 hasConcept C131584629 @default.
W4385698947 hasConcept C148762608 @default.
W4385698947 hasConcept C15744967 @default.
W4385698947 hasConcept C169760540 @default.
W4385698947 hasConcept C41008148 @default.
W4385698947 hasConcept C48103436 @default.
W4385698947 hasConcept C78519656 @default.
W4385698947 hasConceptScore W4385698947C11413529 @default.
W4385698947 hasConceptScore W4385698947C121332964 @default.
W4385698947 hasConceptScore W4385698947C127413603 @default.
W4385698947 hasConceptScore W4385698947C131584629 @default.
W4385698947 hasConceptScore W4385698947C148762608 @default.
W4385698947 hasConceptScore W4385698947C15744967 @default.
W4385698947 hasConceptScore W4385698947C169760540 @default.
W4385698947 hasConceptScore W4385698947C41008148 @default.
W4385698947 hasConceptScore W4385698947C48103436 @default.
W4385698947 hasConceptScore W4385698947C78519656 @default.
W4385698947 hasLocation W43856989471 @default.
W4385698947 hasOpenAccess W4385698947 @default.
W4385698947 hasPrimaryLocation W43856989471 @default.
W4385698947 hasRelatedWork W1520323361 @default.
W4385698947 hasRelatedWork W1578462026 @default.
W4385698947 hasRelatedWork W2001328720 @default.
W4385698947 hasRelatedWork W2055092218 @default.
W4385698947 hasRelatedWork W2101607104 @default.
W4385698947 hasRelatedWork W2112650290 @default.
W4385698947 hasRelatedWork W2138730144 @default.
W4385698947 hasRelatedWork W2412133698 @default.
W4385698947 hasRelatedWork W3033872672 @default.
W4385698947 hasRelatedWork W4321465944 @default.
W4385698947 isParatext "false" @default.
W4385698947 isRetracted "false" @default.
W4385698947 workType "peer-review" @default.