Structure-based dissociations provide agnostic evidence to the multiple-systems debate
Related Articles
MARBLE: interpretable representations of neural population dynamics using geometric deep learning
The dynamics of neuron populations commonly evolve on low-dimensional manifolds. Thus, we need methods that learn the dynamical processes over neural manifolds to infer interpretable and consistent latent representations. We introduce a representation learning method, MARBLE, which decomposes on-manifold dynamics into local flow fields and maps them into a common latent space using unsupervised geometric deep learning. In simulated nonlinear dynamical systems, recurrent neural networks and experimental single-neuron recordings from primates and rodents, we discover emergent low-dimensional latent representations that parametrize high-dimensional neural dynamics during gain modulation, decision-making and changes in the internal state. These representations are consistent across neural networks and animals, enabling the robust comparison of cognitive computations. Extensive benchmarking demonstrates state-of-the-art within- and across-animal decoding accuracy of MARBLE compared to current representation learning approaches, with minimal user input. Our results suggest that a manifold structure provides a powerful inductive bias to develop decoding algorithms and assimilate data across experiments.
Surface-hydroxylated single-atom catalyst with an isolated Co-O-Zn configuration achieves high selectivity in regulating active species
Single-atom catalysts (SACs) are emerging as potent tools for the selective regulation of active species, offering substantial promise for green and sustainable Fenton catalysis. However, current SACs face limitations due to the specificity of their supports, which only allow selective regulation within certain oxidant systems. This constraint makes targeted regulation across different systems challenging. In response, this study designs a SAC, termed CoSAs-ZnO, featuring surface hydroxylation and an isolated asymmetric Co-O-Zn configuration. This SAC can realize a nearly 100% selective generation of sulfate radicals (SO4•−) and singlet oxygen (1O2) in peroxymonosulfate (PMS) and peracetic acid (PAA) systems, respectively. Moreover, the PMS-activated system can efficiently treat electron-deficient-dominated and refractory benzoic acid wastewater, achieving 100.0% removal in multiple consecutive pilot-scale experiments. The PAA-activated system facilitates the rapid conversion of benzyl alcohol to benzaldehyde, with a high selectivity of 89.0%. Detailed DFT calculations reveal that the surface hydroxyl groups on ZnO play a critical role in modulating the adsorption configurations of the oxidants, thus enabling the selective generation of specific active species in each system. This study provides insights into the design of SACs for multifunctional applications and paves the way for their deployment in wastewater treatment and high-value chemical conversion.
Human-structure and human-structure-human interaction in electro-quasistatic regime
Augmented living equipped with electronic devices requires widespread connectivity and a low-loss communication medium for humans to interact with ambient technologies. However, traditional radiative radio frequency-based communications require wireless pairing to ensure specificity during information exchange, and with their broadcasting nature, these incur energy absorption from the surroundings. Recent advancements in electroquasistatic body-coupled communication have shown great promise by utilizing conductive objects like the human body as a communication medium. Here we propose a fundamental set of modalities of non-radiative interaction by guiding electroquasistatic signals through conductive structures between humans and surrounding electronic devices. Our approach offers pairing-free communication specificity and lower path loss during touch. Here, we propose two modalities: Human-Structure Interaction and Human-Structure Human Interaction with wearable devices. We validate our theoretical understanding with numerical electromagnetic simulations and experiments to show the feasibility of the proposed approach. A demonstration of the real-time transfer of an audio signal employing an human body communications-based Human-Structure Interaction link is presented to highlight the practical impact of this work. The proposed techniques can potentially influence Human-Machine Interaction research, including the development of assistive technology for augmented living and personalized healthcare.
T-cell receptor structures and predictive models reveal comparable alpha and beta chain structural diversity despite differing genetic complexity
T-cell receptor (TCR) structures are currently under-utilised in early-stage drug discovery and repertoire-scale informatics. Here, we leverage a large dataset of solved TCR structures from Immunocore to evaluate the current state-of-the-art for TCR structure prediction, and identify which regions of the TCR remain challenging to model. Through clustering analyses and the training of a TCR-specific model capable of large-scale structure prediction, we find that the alpha chain VJ-recombined loop (CDR3α) is as structurally diverse and correspondingly difficult to predict as the beta chain VDJ-recombined loop (CDR3β). This differentiates TCR variable domain loops from the genetically analogous antibody loops and supports the conjecture that both TCR alpha and beta chains are deterministic of antigen specificity. We hypothesise that the larger number of alpha chain joining genes compared to beta chain joining genes compensates for the lack of a diversity gene segment. We also provide over 1.5M predicted TCR structures to enable repertoire structural analysis and elucidate strategies towards improving the accuracy of future TCR structure predictors. Our observations reinforce the importance of paired TCR sequence information and capture the current state-of-the-art for TCR structure prediction, while our model and 1.5M structure predictions enable the use of structural TCR information at an unprecedented scale.
Error-driven upregulation of memory representations
Learning an association does not always succeed on the first attempt. Previous studies associated increased error signals in posterior medial frontal cortex with improved memory formation. However, the neurophysiological mechanisms that facilitate post-error learning remain poorly understood. To address this gap, participants performed a feedback-based association learning task and a 1-back localizer task. Increased hemodynamic responses in posterior medial frontal cortex were found for internal and external origins of memory error evidence, and during post-error encoding success as quantified by subsequent recall of face-associated memories. A localizer-based machine learning model displayed a network of cognitive control regions, including posterior medial frontal and dorsolateral prefrontal cortices, whose activity was related to face-processing evidence in the fusiform face area. Representation strength was higher during failed recall and increased during encoding when subsequent recall succeeded. These data enhance our understanding of the neurophysiological mechanisms of adaptive learning by linking the need for learning with increased processing of the relevant stimulus category.
Responses