Information processing in the brain depends upon the interactions of diverse signaling units, neurons, which vary in their circuit positions, biophysical and biochemical properties, and electrophysiological phenotypes. These distinguishing properties enable neurons to serve different computational and functional roles. For instance, the fast-spiking inhibitory subclasss, which synchronizes its firing at fast frequencies and makes divergent and convergent contacts onto the persimomatic compartments of its targets, is hypothesized to play an important role in regulating the timing and probability of spike output and thus the flow of information through the brain (reviewed in Chapter 1). However, the functional consequences of synchronized inhibition will depend on the integrative properties and circuit locations of the recipient neurons. Using a combination of in-vitro whole cell electrophysiology, modeling, and model-neuron hybrids (dynamic clamp), we determined how fast synchronized inhibition interacts with integrative properties of a variety of neuronal types to regulate the rate (Chapter 2) and timing (Chapter 3) of spike generation. We find that a neuron's intrinsic physiology substantially affects the ability of fast synchronized inhibition to control neuronal responsiveness. In addition, we demonstrate how the relevant physiology can be flexibly altered by contextual and neuromodulatory factors. The results of these experiments suggest that synchronized fast-spiking activity can differentially affect various circuit elements and each element's responsiveness can be adjusted on a range of time scales to suit the cortex's changing computational requirements. Because the brain is one of the most energetically expensive organs in the body with action potential generation accounting for significant portion of the energy usage, we hypothesized that differences in neuronal properties may also serve to minimize energy consumption subject to functional constraints. Again using electrophysiology, modeling, and dynamic clamp, we compared the energy needed to produce action potentials singly and in trains for a wide range of channel densities and kinetic parameters, and examined which combinations of parameters maximized spiking function while minimizing energetic cost (Chapter 4). We found evidence supporting our hypothesis in a wide range of neurons from several species. We conclude that neuronal biophysics are tuned to perform cost-effective functions

We demonstrate that inference-based goal-directed behavior can be done by utilizing the temporal gradients in re-current neural network (RNN). The RNN learns a dynamic sensorimotor forward model. Once the RNN is trained, it can beused to execute active-inference-based, goal-directed policy optimization. The internal neural activities of the trained RNNessentially model the predictive state of the controlled entity. The implemented optimization process projects the neural activ-ities into the future via the RNN recurrences following a tentative sequence of motor commands (encoded in neurons akin torecurrent parametric biases). This sequence is adapted by back-projecting the error between the forward-projected hypotheticalstates and desired (goal-like) system states onto the motor commands. Few cycles of forward projection and goal-based errorbackpropagation yield the sequences of motor commands that control the dynamical systems. As an example, we show that atrained RNN model can be used to effectively control a quadrocopter-like system.

Purpose of review

The human circadian system creates and maintains cellular and systemic rhythmicity essential to homeostasis. Loss of circadian rhythmicity fundamentally affects the neuroendocrine, immune and autonomic system, similar to chronic stress and, thus, may play a central role in the development of stress-related disorders. This article focuses on the role of circadian misalignment in the pathophysiology of posttraumatic stress disorder (PTSD).

Recent findings

Sleep disruption is a core feature of PTSD supporting the important supraordinate pathophysiological role of circadian system in PTSD. Furthermore, direct and indirect human and animal PTSD research suggests circadian system linked neuroendocrine, immune, metabolic and autonomic dysregulation with blunted diurnal rhythms, specific sleep pattern pathologies and cognitive deficits, as well as endocannabinoid and neuropeptide Y system alterations and altered circadian gene expression, linking circadian misalignment to PTSD pathophysiology.

Summary

PTSD development is associated with chronodisruption findings. Evaluation and treatment of sleep and circadian disruption should be the first steps in PTSD management. State-of-the-art methods of circadian rhythm assessment should be applied to bridge the gap between clinical significance and limited understanding of the relationship between traumatic stress, sleep and circadian system.

Character rigs are procedural systems that compute the shape of an animated character for a given pose. They can be highly complex and must account for bulges, wrinkles, and other aspects of a character's appearance. When comparing film-quality character rigs with those designed for real-time applications, there is typically a substantial and readily apparent difference in the quality of the mesh deformations. Real-time rigs are limited by a computational budget and often trade realism for performance. Rigs for film do not have this same limitation, and character riggers can make the rig as complicated as necessary to achieve realistic deformations. However, increasing the rig complexity slows rig evaluation, and the animators working with it can become less efficient and may experience frustration. In this paper, we present a method to reduce the time required to compute mesh deformations for film-quality rigs, allowing better interactivity during animation authoring and use in real-time games and applications. Our approach learns the deformations from an existing rig by splitting the mesh deformation into linear and nonlinear portions. The linear deformations are computed directly from the transformations of the rig's underlying skeleton. We use deep learning methods to approximate the remaining nonlinear portion. In the examples we show from production rigs used to animate lead characters, our approach reduces the computational time spent on evaluating deformations by a factor of 5×-10×. This significant savings allows us to run the complex, film-quality rigs in real-time even when using a CPU-only implementation on a mobile device.

Humans learn that temporarily occluded objects continue to exist within the first months of their lives. Deep learning mod- els, on the other hand, struggle to generalize such concepts from observations, due to missing proper inductive biases. Here, we introduce the first self-supervised interpretable ma- chine learning model that learns about object permanence di- rectly from video data without supervision. We augment a slot- based autoregressive deep learning system with the ability to adaptively and selectively fuse latent imaginations with pixel- based observations into consistent object-specific ‘what' and ‘where' encodings over time. We show that (i) Loci-Looped tracks objects through occlusions and anticipates their reap- pearance while outperforming state-of-the-art baseline models, (ii) Loci-Looped shows signs of surprise when the principle of object permanence is violated, and (iii) Loci-Looped's internal latent loop is key for learning object permanence.

Motivated by the close relation of predictive coding and activeinference to cognition, we introduce a dynamic artificial neu-ral network-based (ANN) adaptation process, which we termREPRISE: REtrospective and PRospective Inference SchEme.REPRISE first executes a retrospective inference process, in-ferring the unobservable contextual state that best explains itsrecently encountered sensorimotor experiences. It then exe-cutes a prospective inference process, inferring upcoming mo-tor activities in the light of the inferred contextual state anda given goal state. First, the ANN – a recurrent neural net-work – is trained to learn one sensorimotor temporal forwardmodel, that is, the sensorimotor contingencies generated by thebehavior of three moving or flying vehicles. During training,additional three bits are provided as input, indicating whichmode currently applies. After training, goal-directed controland system state inference are activated: Given a goal state,the system imagines a motor command sequence optimizing itwith the prospective objective to minimize the distance to thegoal. Meanwhile, the system evaluates the encountered sen-sorimotor contingencies retrospectively, adapting its vehicleestimation activities and, in order to maintain coherence, theneural hidden states accordingly. This ANN’s ’mind’ is thuscontinuously imagining the future and reflecting on the past –showing superior performance on the posed control problems.The architecture effectively demonstrates that neural error sig-nals and neural activities can be projected into the past and intothe future, respectively, optimizing both neural context codesthat approximately generate the recent past and upcoming be-havior in the light of desired goal states.

Spliceosome formation is initiated by the recognition of the 5′ splice site through formation of an RNA duplex between the 5′ splice site and U1 snRNA. We have previously shown that RNA duplex formation between U1 snRNA and the 5′ splice site can protect pre-mRNAs from degradation prior to splicing. This initial RNA duplex must be disrupted to expose the 5′ splice site sequence for base pairing with U6 snRNA and to form the active spliceosome. Here, we investigated whether hyperstabilization of the U1 snRNA/5′ splice site duplex interferes with splicing efficiency in human cell lines or nuclear extracts. Unlike observations in Saccharomyces cerevisiae, we demonstrate that an extended U1 snRNA/5′ splice site interaction does not decrease splicing efficiency, but rather increases 5′ splice site recognition and exon inclusion. However, low complementarity of the 5′ splice site to U1 snRNA significantly increases exon skipping and RNA degradation. Although the splicing mechanisms are conserved between human and S.cerevisiae, these results demonstrate that distinct differences exist in the activation of the spliceosome.

Hopelessness is associated with mortality in patients with cardiac disease even after accounting for severity of depression. We sought to determine whether a polymorphism in the promoter region of the serotonin transporter gene (5-HTTLPR) is associated with increased hopelessness, and whether this effect is modified by sex, age, antidepressant use or depression in patients with coronary heart disease. We conducted a cross-sectional study of 870 patients with stable coronary heart disease. Our primary outcomes were hopelessness score (range 0-8) and hopeless category (low, moderate and high) as measured by the Everson hopelessness scale. Analysis of covariance and ordinal logistic regression were used to examine the independent association of genotype with hopelessness. Compared to patients with l/l genotype, adjusted odds of a higher hopeless category increased by 35% for the l/s genotype and 80% for s/s genotype (p-value for trend = 0.004). Analysis of covariance demonstrated that the effect of 5-HTTLPR genotype on hopelessness was modified by sex (.04), but not by racial group (p = 0.63). Among men, odds of higher hopeless category increased by 40% for the l/s genotype and by 2.3-fold for s/s genotype (p-value p < 0.001), compared to no effect in the smaller female sample (p = 0.42). Results stratified by race demonstrated a similar dose-response effect of the s allele on hopelessness across racial groups. We found that the 5-HTTLPR is independently associated with hopelessness among men with cardiovascular disease.