Dendritic Integration of Sensory Evidence in Perceptual Decision-Making.

Perceptual decisions require the accumulation of sensory information to a response criterion. Most accounts of how the brain performs this process of temporal integration have focused on evolving patterns of spiking activity. We report that subthreshold changes in membrane voltage can represent accumulating evidence before a choice. αß core Kenyon cells (αßc KCs) in the mushroom bodies of fruit flies integrate odor-evoked synaptic inputs to action potential threshold at timescales matching the speed of olfactory discrimination. The forkhead box P transcription factor (FoxP) sets neuronal integration and behavioral decision times by controlling the abundance of the voltage-gated potassium channel Shal (KV4) in αßc KC dendrites. αßc KCs thus tailor, through a particular constellation of biophysical properties, the generic process of synaptic integration to the demands of sequential sampling.

stDiff: a diffusion model for imputing spatial transcriptomics through single-cell transcriptomics.

Spatial transcriptomics (ST) has become a powerful tool for exploring the spatial organization of gene expression in tissues. Imaging-based methods, though offering superior spatial resolutions at the single-cell level, are limited in either the number of imaged genes or the sensitivity of gene detection. Existing approaches for enhancing ST rely on the similarity between ST cells and reference single-cell RNA sequencing (scRNA-seq) cells. In contrast, we introduce stDiff, which leverages relationships between gene expression abundance in scRNA-seq data to enhance ST. stDiff employs a conditional diffusion model, capturing gene expression abundance relationships in scRNA-seq data through two Markov processes: one introducing noise to transcriptomics data and the other denoising to recover them. The missing portion of ST is predicted by incorporating the original ST data into the denoising process. In our comprehensive performance evaluation across 16 datasets, utilizing multiple clustering and similarity metrics, stDiff stands out for its exceptional ability to preserve topological structures among cells, positioning itself as a robust solution for cell population identification. Moreover, stDiff's enhancement outcomes closely mirror the actual ST data within the batch space. Across diverse spatial expression patterns, our model accurately reconstructs them, delineating distinct spatial boundaries. This highlights stDiff's capability to unify the observed and predicted segments of ST data for subsequent analysis. We anticipate that stDiff, with its innovative approach, will contribute to advancing ST imputation methodologies.

Guided diffusion for molecular generation with interaction prompt.

Molecular generative models have exhibited promising capabilities in designing molecules from scratch with high binding affinities in a predetermined protein pocket, offering potential synergies with traditional structural-based drug design strategy. However, the generative processes of such models are random and the atomic interaction information between ligand and protein are ignored. On the other hand, the ligand has high propensity to bind with residues called hotspots. Hotspot residues contribute to the majority of the binding free energies and have been recognized as appealing targets for designed molecules. In this work, we develop an interaction prompt guided diffusion model, InterDiff to deal with the challenges. Four kinds of atomic interactions are involved in our model and represented as learnable vector embeddings. These embeddings serve as conditions for individual residue to guide the molecular generative process. Comprehensive in silico experiments evince that our model could generate molecules with desired ligand-protein interactions in a guidable way. Furthermore, we validate InterDiff on two realistic protein-based therapeutic agents. Results show that InterDiff could generate molecules with better or similar binding mode compared to known targeted drugs.

HiCDiff: single-cell Hi-C data denoising with diffusion models.

The genome-wide single-cell chromosome conformation capture technique, i.e. single-cell Hi-C (ScHi-C), was recently developed to interrogate the conformation of the genome of individual cells. However, single-cell Hi-C data are much sparser than bulk Hi-C data of a population of cells, and noise in single-cell Hi-C makes it difficult to apply and analyze them in biological research. Here, we developed the first generative diffusion models (HiCDiff) to denoise single-cell Hi-C data in the form of chromosomal contact matrices. HiCDiff uses a deep residual network to remove the noise in the reverse process of diffusion and can be trained in both unsupervised and supervised learning modes. Benchmarked on several single-cell Hi-C test datasets, the diffusion models substantially remove the noise in single-cell Hi-C data. The unsupervised HiCDiff outperforms most supervised non-diffusion deep learning methods and achieves the performance comparable to the state-of-the-art supervised deep learning method in terms of multiple metrics, demonstrating that diffusion models are a useful approach to denoising single-cell Hi-C data. Moreover, its good performance holds on denoising bulk Hi-C data.

RNA polymerase sliding on DNA can couple the transcription of nearby bacterial operons.

DNA transcription initiates after an RNA polymerase (RNAP) molecule binds to the promoter of a gene. In bacteria, the canonical picture is that RNAP comes from the cytoplasmic pool of freely diffusing RNAP molecules. Recent experiments suggest the possible existence of a separate pool of polymerases, competent for initiation, which freely slide on the DNA after having terminated one round of transcription. Promoter-dependent transcription reinitiation from this pool of posttermination RNAP may lead to coupled initiation at nearby operons, but it is unclear whether this can occur over the distance and timescales needed for it to function widely on a bacterial genome in vivo. Here, we mathematically model the hypothesized reinitiation mechanism as a diffusion-to-capture process and compute the distances over which significant interoperon coupling can occur and the time required. These quantities depend on molecular association and dissociation rate constants between DNA, RNAP, and the transcription initiation factor σ70; we measure these rate constants using single-molecule experiments in vitro. Our combined theory/experimental results demonstrate that efficient coupling can occur at physiologically relevant σ70 concentrations and on timescales appropriate for transcript synthesis. Coupling is efficient over terminator-promoter distances up to ∼1,000 bp, which includes the majority of terminator-promoter nearest neighbor pairs in the Escherichia coli genome. The results suggest a generalized mechanism that couples the transcription of nearby operons and breaks the paradigm that each binding of RNAP to DNA can produce at most one messenger RNA.

Stochastic Turing patterns of trichomes in Arabidopsis leaves.

Biological patterns that emerge during the morphogenesis of multicellular organisms can display high precision at large scales, while at cellular scales, cells exhibit large fluctuations stemming from cell-cell differences in molecular copy numbers also called demographic noise. We study the conflicting interplay between high precision and demographic noise in trichome patterns on the epidermis of wild-type Arabidopsis thaliana leaves, as a two-dimensional model system. We carry out a statistical characterization of these patterns and show that their power spectra display fat tails-a signature compatible with noise-driven stochastic Turing patterns-which are absent in power spectra of patterns driven by deterministic instabilities. We then present a theoretical model that includes demographic noise stemming from birth-death processes of genetic regulators which we study analytically and by stochastic simulations. The model captures the observed experimental features of trichome patterns.

Cortical ß Power Reflects a Neural Implementation of Decision Boundary Collapse in Speeded Decisions.

A prominent account of decision-making assumes that information is accumulated until a fixed response threshold is crossed. However, many decisions require weighting of information appropriately against time. Collapsing response thresholds are a mathematically optimal solution to this decision problem. However, our understanding of the neurocomputational mechanisms underlying dynamic response thresholds remains significantly incomplete. To investigate this issue, we used a multistage drift-diffusion model (DDM) and also analyzed EEG ß power lateralization (BPL). The latter served as a neural proxy for decision signals. We analyzed a large dataset (n = 863; 434 females and 429 males) from a speeded flanker task and data from an independent confirmation sample (n = 119; 70 females and 49 males). We showed that a DDM with collapsing decision thresholds, a process wherein the decision boundary reduces over time, captured participants' time-dependent decision policy more accurately than a model with fixed thresholds. Previous research suggests that BPL over motor cortices reflects features of a decision signal and that its peak, coinciding with the motor response, may serve as a neural proxy for the decision threshold. We show that BPL around the response decreased with increasing RTs. Together, our findings offer compelling evidence for the existence of collapsing decision thresholds in decision-making processes.

Empathy incites a stable prosocial decision bias.

Empathy toward suffering individuals serves as potent driver for prosocial behavior. However, it remains unclear whether prosociality induced by empathy for another person's pain persists once that person's suffering diminishes. To test this, participants underwent functional magnetic resonance imaging while performing a binary social decision task that involved allocation of points to themselves and another person. In block one, participants completed the task after witnessing frequent painful stimulation of the other person, and in block two, after observing low frequency of painful stimulation. Drift-diffusion modeling revealed an increased initial bias toward making prosocial decisions in the first block compared with baseline that persisted in the second block. These results were replicated in an independent behavioral study. An additional control study showed that this effect may be specific to empathy as stability was not evident when prosocial decisions were driven by a social norm such as reciprocity. Increased neural activation in dorsomedial prefrontal cortex was linked to empathic concern after witnessing frequent pain and to a general prosocial decision bias after witnessing rare pain. Altogether, our findings show that empathy for pain elicits a stable inclination toward making prosocial decisions even as their suffering diminishes.

High-value decisions are fast and accurate, inconsistent with diminishing value sensitivity.

It is a widely held belief that people's choices are less sensitive to changes in value as value increases. For example, the subjective difference between $11 and $12 is believed to be smaller than between $1 and $2. This idea is consistent with applications of the Weber-Fechner Law and divisive normalization to value-based choice and with psychological interpretations of diminishing marginal utility. According to random utility theory in economics, smaller subjective differences predict less accurate choices. Meanwhile, in the context of sequential sampling models in psychology, smaller subjective differences also predict longer response times. Based on these models, we would predict decisions between high-value options to be slower and less accurate. In contrast, some have argued on normative grounds that choices between high-value options should be made with less caution, leading to faster and less accurate choices. Here, we model the dynamics of the choice process across three different choice domains, accounting for both discriminability and response caution. Contrary to predictions, we mostly observe faster and more accurate decisions (i.e., higher drift rates) between high-value options. We also observe that when participants are alerted about incoming high-value decisions, they exert more caution and not less. We rule out several explanations for these results, using tasks with both subjective and objective values. These results cast doubt on the notion that increasing value reduces discriminability.

Equivariant score-based generative diffusion framework for 3D molecules.

BACKGROUND: Molecular biology is crucial for drug discovery, protein design, and human health. Due to the vastness of the drug-like chemical space, depending on biomedical experts to manually design molecules is exceedingly expensive. Utilizing generative methods with deep learning technology offers an effective approach to streamline the search space for molecular design and save costs. This paper introduces a novel E(3)-equivariant score-based diffusion framework for 3D molecular generation via SDEs, aiming to address the constraints of unified Gaussian diffusion methods. Within the proposed framework EMDS, the complete diffusion is decomposed into separate diffusion processes for distinct components of the molecular feature space, while the modeling processes also capture the complex dependency among these components. Moreover, angle and torsion angle information is integrated into the networks to enhance the modeling of atom coordinates and utilize spatial information more effectively. RESULTS: Experiments on the widely utilized QM9 dataset demonstrate that our proposed framework significantly outperforms the state-of-the-art methods in all evaluation metrics for 3D molecular generation. Additionally, ablation experiments are conducted to highlight the contribution of key components in our framework, demonstrating the effectiveness of the proposed framework and the performance improvements of incorporating angle and torsion angle information for molecular generation. Finally, the comparative results of distribution show that our method is highly effective in generating molecules that closely resemble the actual scenario. CONCLUSION: Through the experiments and comparative results, our framework clearly outperforms previous 3D molecular generation methods, exhibiting significantly better capacity for modeling chemically realistic molecules. The excellent performance of EMDS in 3D molecular generation brings novel and encouraging opportunities for tackling challenging biomedical molecule and protein scenarios.

Two fundamentally different mechanisms by which unconscious information impairs behavioral performance: Evidence from fMRI and computational modeling.

It is increasingly clear that unconscious information impairs the performance of the corresponding action when the instruction to act is delayed. However, whether this impairment occurs at the response level or at the perceptual level remains controversial. This study used fMRI and a computational model with a pre-post design to address this elusive issue. The fMRI results showed that when the unconscious information containing strong stimulus-response associations was irrelevant to subsequent stimuli, the precuneus in the parietal lobe, which is thought to be involved in sensorimotor processing, was activated. In contrast, when the unconscious information was relevant to subsequent stimuli, regardless of the strength of the stimulus-response associations, some regions in the occipital and temporal cortices, which are thought to be involved in visual perceptual processing, were activated. In addition, the percent signal change in the regions of interest associated with motor inhibition was modulated by compatibility in the irrelevant but not in the relevant stimuli conditions. Modeling of behavioral data further supported that the irrelevant and relevant stimuli conditions involved fundamentally different mechanisms. Our finding reconciles the debate about the mechanism by which unconscious information impairs action performance and has important implications for understanding of unconscious cognition.

Deep latent variable joint cognitive modeling of neural signals and human behavior.

As the field of computational cognitive neuroscience continues to expand and generate new theories, there is a growing need for more advanced methods to test the hypothesis of brain-behavior relationships. Recent progress in Bayesian cognitive modeling has enabled the combination of neural and behavioral models into a single unifying framework. However, these approaches require manual feature extraction, and lack the capability to discover previously unknown neural features in more complex data. Consequently, this would hinder the expressiveness of the models. To address these challenges, we propose a Neurocognitive Variational Autoencoder (NCVA) to conjoin high-dimensional EEG with a cognitive model in both generative and predictive modeling analyses. Importantly, our NCVA enables both the prediction of EEG signals given behavioral data and the estimation of cognitive model parameters from EEG signals. This novel approach can allow for a more comprehensive understanding of the triplet relationship between behavior, brain activity, and cognitive processes.

Manipulating Prior Beliefs Causally Induces Under- and Overconfidence.

Humans differ vastly in the confidence they assign to decisions. Although such under- and overconfidence relate to fundamental life outcomes, a computational account specifying the underlying mechanisms is currently lacking. We propose that prior beliefs in the ability to perform a task explain confidence differences across participants and tasks, despite similar performance. In two perceptual decision-making experiments, we show that manipulating prior beliefs about performance during training causally influences confidence in healthy adults (N = 50 each; Experiment 1: 8 men, one nonbinary; Experiment 2: 5 men) during a test phase, despite unaffected objective performance. This is true when prior beliefs are induced via manipulated comparative feedback and via manipulated training-phase difficulty. Our results were accounted for within an accumulation-to-bound model, explicitly modeling prior beliefs on the basis of earlier task exposure. Decision confidence is quantified as the probability of being correct conditional on prior beliefs, causing under- or overconfidence. We provide a fundamental mechanistic insight into the computations underlying under- and overconfidence.

A mathematical model of signalling molecule-mediated processes during regeneration of osteochondral defects after chondrocyte implantation.

Treating bone-cartilage defects is a fundamental clinical problem. The ability of damaged cartilage to self-repair is limited due to its avascularity. Left untreated, these defects can lead to osteoarthritis. Details of osteochondral defect repair are elusive, but animal models indicate healing occurs via an endochondral ossification-like process, similar to that in the growth plate. In the growth plate, the signalling molecules parathyroid hormone-related protein (PTHrP) and Indian Hedgehog (Ihh) form a feedback loop regulating chondrocyte hypertrophy, with Ihh inducing and PTHrP suppressing hypertrophy. To better understand this repair process and to explore the regulatory role of signalling molecules on the regeneration process, we formulate a reaction-diffusion mathematical model of osteochondral defect regeneration after chondrocyte implantation. The drivers of healing are assumed to be chondrocytes and osteoblasts, and their interaction via signalling molecules. We model cell proliferation, migration and chondrocyte hypertrophy, and matrix production and conversion, spatially and temporally. We further model nutrient and signalling molecule diffusion and their interaction with the cells. We consider the PTHrP-Ihh feedback loop as the backbone mechanisms but the model is flexible to incorporate extra signalling mechanisms if needed. Our mathematical model is able to represent repair of osteochondral defects, starting with cartilage formation throughout the defect. This is followed by chondrocyte hypertrophy, matrix calcification and bone formation deep inside the defect, while cartilage at the surface is maintained and eventually separated from the deeper bone by a thin layer of calcified cartilage. The complete process requires around 48 months. A key highlight of the model demonstrates that the PTHrP-Ihh loop alone is insufficient and an extra mechanism is required to initiate chondrocyte hypertrophy, represented by a critical cartilage density. A parameter sensitivity study reveals that the timing of the repair process crucially depends on parameters, such as the critical cartilage density, and those describing the actions of PTHrP to suppress hypertrophy, such as its diffusion coefficient, threshold concentration and degradation rate.

How do productivity gradient and diffusion shape patterns in a plant-herbivore grazing system?

Natural systems show heterogeneous patchy distributions of vegetation over large landscapes. Reaction-diffusion systems can demonstrate such heterogeneity of species distributions. Here, we analyse a reaction-diffusion model of plant-herbivore interactions in two-dimensional space to illustrate non-homogeneous distributions of plants and herbivores. The non-spatial system shows bottom-up control, where herbivore density is low under low and high primary productivity but increased at intermediate productivity. In addition, the non-spatial system provides bistability between a dense vegetation state devoid of herbivores and a coexisting state of plants and herbivores. In the spatiotemporal model, we give analytical conditions of occurring diffusion-driven (Turing) instability, where a novel point in our model is the relative dispersal of herbivores, which represents the movement of herbivores from a higher to a lower vegetation state in addition to the self-diffusion of both species. It is shown that heterogeneity in the population distribution does not occur if the relative dispersal of herbivores is low, but it appears in the opposite case. Due to bistability in the underlying non-spatial system, the spatiotemporal model produces initial value-dependent patterns. The two initial values make different patterns despite having the same primary productivity and relative dispersal rate. As productivity increases with a given relative herbivore dispersal, pattern transition occurs from a blend of stripes and spots of low vegetation state to a predominantly low-density vegetation state with smaller patches of densely vegetated states with one initial value. On the contrary, a discernible change in vegetation patterns from cold spots in the dense vegetation to hot stripes in the primarily low-vegetated state is noticed under the other initial population value. Furthermore, the population distributions of plants and herbivores in the entire domain after a long period are heterogeneous for both initial values, provided the relative herbivore dispersal is substantial. We estimated mean population densities to observe species fitness in the whole domain under variable productivity. When productivity is high, the mean population density of plants may go up or down, depending on the herbivore's relative dispersal rate. In contrast to the bottom-up control dynamics of the non-spatial system, the system exhibits a top-down control under high relative dispersal, where the herbivore regulates vegetation growth under high productivity. On the other hand, herbivores are extinct under high productivity if the relative dispersal is low.

Teleological reasoning bias is predicted by pupil dynamics: Evidence for the extensive integration account of bias in reasoning.

Teleological reasoning is the tendency for humans to see purpose and intentionality in natural phenomena when there is none. In this study, we assess three competing theories on how bias in reasoning arises by examining performance on a teleological reasoning task while measuring pupil size and response times. We replicate that humans (N = 45) are prone to accept false teleological explanations. Further, we show that errors on the teleological reasoning task are associated with slower response times, smaller baseline pupil size, and larger pupil dilations. The results are in line with the single-process extensive integration account and directly oppose predictions from dual-processing accounts. Lastly, by modeling responses with a drift-diffusion model, we find that larger baseline pupil size is associated with lower decision threshold and higher drift rate, whereas larger pupil dilations are associated with higher decision threshold and lower drift rate. The results highlight the role of neural gain and the Locus Coeruleus-Norepinephrine system in modulating evidence integration and bias in reasoning. Thus, teleological reasoning and susceptibility to bias likely arise due to extensive processing rather than through fast and effortless processing.

Attention facilitates initiation of perceptual decision making: a combined psychophysical and electroencephalography study.

Humans can selectively process information and make decisions by directing their attention to desired locations in their daily lives. Numerous studies have shown that attention increases the rate of correct responses and shortens reaction time, and it has been hypothesized that this phenomenon is caused by an increase in sensitivity of the sensory signals to which attention is directed. The present study employed psychophysical methods and electroencephalography (EEG) to test the hypothesis that attention accelerates the onset of information accumulation. Participants were asked to discriminate the motion direction of one of two random dot kinematograms presented on the left and right sides of the visual field, one of which was cued by an arrow in 80% of the trials. The drift-diffusion model was applied to the percentage of correct responses and reaction times in the attended and unattended fields of view. Attention primarily increased sensory sensitivity and shortened the time unrelated to decision making. Next, we measured centroparietal positivity (CPP), an EEG measure associated with decision making, and found that CPP latency was shorter in attended trials than in unattended trials. These results suggest that attention not only increases sensory sensitivity but also accelerates the initiation of decision making.

Diverse effects of gaze direction on heading perception in humans.

Gaze change can misalign spatial reference frames encoding visual and vestibular signals in cortex, which may affect the heading discrimination. Here, by systematically manipulating the eye-in-head and head-on-body positions to change the gaze direction of subjects, the performance of heading discrimination was tested with visual, vestibular, and combined stimuli in a reaction-time task in which the reaction time is under the control of subjects. We found the gaze change induced substantial biases in perceived heading, increased the threshold of discrimination and reaction time of subjects in all stimulus conditions. For the visual stimulus, the gaze effects were induced by changing the eye-in-world position, and the perceived heading was biased in the opposite direction of gaze. In contrast, the vestibular gaze effects were induced by changing the eye-in-head position, and the perceived heading was biased in the same direction of gaze. Although the bias was reduced when the visual and vestibular stimuli were combined, integration of the 2 signals substantially deviated from predictions of an extended diffusion model that accumulates evidence optimally over time and across sensory modalities. These findings reveal diverse gaze effects on the heading discrimination and emphasize that the transformation of spatial reference frames may underlie the effects.

Midfrontal theta phase underlies evidence accumulation and response thresholding in cognitive control.

Cognitive control involves evidence accumulation and response thresholding, but the neural underpinnings of these 2 processes are poorly understood. Based on recent findings that midfrontal theta phase coordinates the correlation between theta power and reaction time during cognitive control, this study investigated whether and how theta phase would modulate the relationships between theta power and evidence accumulation and response thresholding in human participants when they performed a flanker task. Our results confirmed the modulation of theta phase on the correlations between ongoing midfrontal theta power and reaction time under both conditions. Using hierarchical drift-diffusion regression modeling, we found that in both conditions, theta power was positively associated with boundary separation in phase bins with optimal power-reaction time correlations, whereas the power-boundary correlation decreased to nonsignificance in phase bins with reduced power-reaction time correlations. In contrast, the power-drift rate correlation was not modulated by theta phase, but by cognitive conflict. Drift rate was positively correlated with theta power for the bottom-up processing in the non-conflict condition, whereas it was negatively correlated with theta power for the top-down control to address conflict. These findings suggest that evidence accumulation is likely to be a phase-coordinated continuous process, whereas thresholding may be a phase-specific transient process.

Quantifying uncertainty: Air quality forecasting based on dynamic spatial-temporal denoising diffusion probabilistic model.

Air pollution constitutes a substantial peril to human health, thereby catalyzing the evolution of an array of air quality prediction models. These models span from mechanistic and statistical strategies to machine learning methodologies. The burgeoning field of deep learning has given rise to a plethora of advanced models, which have demonstrated commendable performance. However, previous investigations have overlooked the salience of quantifying prediction uncertainties and potential future interconnections among air monitoring stations. Moreover, prior research typically utilized static predetermined spatial relationships, neglecting dynamic dependencies. To address these limitations, we propose a model named Dynamic Spatial-Temporal Denoising Diffusion Probabilistic Model (DST-DDPM) for air quality prediction. Our model is underpinned by the renowned denoising diffusion model, aiding us in discerning indeterminacy. In order to encapsulate dynamic patterns, we design a dynamic context encoder to generate dynamic adjacency matrices, whilst maintaining static spatial information. Furthermore, we incorporate a spatial-temporal denoising model to concurrently learn both spatial and temporal dependencies. Authenticating our model's performance using a real-world dataset collected in Beijing, the outcomes indicate that our model eclipses other baseline models in terms of both short-term and long-term predictions by 1.36% and 11.62% respectively. Finally, we conduct a case study to exhibit our model's capacity to quantify uncertainties.