Mixed vine copula flows for flexible modeling of neural dependencies.

Recordings of complex neural population responses provide a unique opportunity for advancing our understanding of neural information processing at multiple scales and improving performance of brain computer interfaces. However, most existing analytical techniques fall short of capturing the complexity of interactions within the concerted population activity. Vine copula-based approaches have shown to be successful at addressing complex high-order dependencies within the population, disentangled from the single-neuron statistics. However, most applications have focused on parametric copulas which bear the risk of misspecifying dependence structures. In order to avoid this risk, we adopted a fully non-parametric approach for the single-neuron margins and copulas by using Neural Spline Flows (NSF). We validated the NSF framework on simulated data of continuous and discrete types with various forms of dependency structures and with different dimensionality. Overall, NSFs performed similarly to existing non-parametric estimators, while allowing for considerably faster and more flexible sampling which also enables faster Monte Carlo estimation of copula entropy. Moreover, our framework was able to capture low and higher order heavy tail dependencies in neuronal responses recorded in the mouse primary visual cortex during a visual learning task while the animal was navigating a virtual reality environment. These findings highlight an often ignored aspect of complexity in coordinated neuronal activity which can be important for understanding and deciphering collective neural dynamics for neurotechnological applications.

Parametric Copula-GP model for analyzing multidimensional neuronal and behavioral relationships.

One of the main goals of current systems neuroscience is to understand how neuronal populations integrate sensory information to inform behavior. However, estimating stimulus or behavioral information that is encoded in high-dimensional neuronal populations is challenging. We propose a method based on parametric copulas which allows modeling joint distributions of neuronal and behavioral variables characterized by different statistics and timescales. To account for temporal or spatial changes in dependencies between variables, we model varying copula parameters by means of Gaussian Processes (GP). We validate the resulting Copula-GP framework on synthetic data and on neuronal and behavioral recordings obtained in awake mice. We show that the use of a parametric description of the high-dimensional dependence structure in our method provides better accuracy in mutual information estimation in higher dimensions compared to other non-parametric methods. Moreover, by quantifying the redundancy between neuronal and behavioral variables, our model exposed the location of the reward zone in an unsupervised manner (i.e., without using any explicit cues about the task structure). These results demonstrate that the Copula-GP framework is particularly useful for the analysis of complex multidimensional relationships between neuronal, sensory and behavioral variables.

Reward Association Enhances Stimulus-Specific Representations in Primary Visual Cortex.

The potential for neuronal representations of external stimuli to be modified by previous experience is critical for efficient sensory processing and improved behavioral outcomes. To investigate how repeated exposure to a visual stimulus affects its representation in mouse primary visual cortex (V1), we performed two-photon calcium imaging of layer 2/3 neurons and assessed responses before, during, and after the presentation of a repetitive stimulus over 5 consecutive days. We found a stimulus-specific enhancement of the neuronal representation of the repetitively presented stimulus when it was associated with a reward. This was observed both after mice actively learned a rewarded task and when the reward was randomly received. Stimulus-specific enhanced representation resulted both from neurons gaining selectivity and from increased response reliability in previously selective neurons. In the absence of reward, there was either no change in stimulus representation or a decreased representation when the stimulus was viewed at a fixed temporal frequency. Pairing a second stimulus with a reward led to a similar enhanced representation and increased discriminability between the equally rewarded stimuli. Single-neuron responses showed that separate subpopulations discriminated between the two rewarded stimuli depending on whether the stimuli were displayed in a virtual environment or viewed on a single screen. We suggest that reward-associated responses enable the generalization of enhanced stimulus representation across these V1 subpopulations. We propose that this dynamic regulation of visual processing based on the behavioral relevance of sensory input ultimately enhances and stabilizes the representation of task-relevant features while suppressing responses to non-relevant stimuli.

Raptor-Mediated Proteasomal Degradation of Deamidated 4E-BP2 Regulates Postnatal Neuronal Translation and NF-κB Activity.

The translation initiation repressor 4E-BP2 is deamidated in the brain on asparagines N99/N102 during early postnatal brain development. This post-translational modification enhances 4E-BP2 association with Raptor, a central component of mTORC1 and alters the kinetics of excitatory synaptic transmission. We show that 4E-BP2 deamidation is neuron specific, occurs in the human brain, and changes 4E-BP2 subcellular localization, but not its disordered structure state. We demonstrate that deamidated 4E-BP2 is ubiquitinated more and degrades faster than the unmodified protein. We find that enhanced deamidated 4E-BP2 degradation is dependent on Raptor binding, concomitant with increased association with a Raptor-CUL4B E3 ubiquitin ligase complex. Deamidated 4E-BP2 stability is promoted by inhibiting mTORC1 or glutamate receptors. We further demonstrate that deamidated 4E-BP2 regulates the translation of a distinct pool of mRNAs linked to cerebral development, mitochondria, and NF-κB activity, and thus may be crucial for postnatal brain development in neurodevelopmental disorders, such as ASD.