Intention Understanding in Autism

When we observe a motor act (e.g. grasping a cup) done by another individual, we extract, according to how the motor act is performed and its context, two types of information: the goal (grasping) and the intention underlying it (e.g. grasping for drinking). Here we examined whether children with autistic spectrum disorder (ASD) are able to understand these two aspects of motor acts. Two experiments were carried out. In the first, one group of high-functioning children with ASD and one of typically developing (TD) children were presented with pictures showing hand-object interactions and asked what the individual was doing and why. In half of the “why” trials the observed grip was congruent with the function of the object (“why-use” trials), in the other half it corresponded to the grip typically used to move that object (“why-place” trials). The results showed that children with ASD have no difficulties in reporting the goals of individual motor acts. In contrast they made several errors in the why task with all errors occurring in the “why-place” trials. In the second experiment the same two groups of children saw pictures showing a hand-grip congruent with the object use, but within a context suggesting either the use of the object or its placement into a container. Here children with ASD performed as TD children, correctly indicating the agent''s intention. In conclusion, our data show that understanding others'' intentions can occur in two ways: by relying on motor information derived from the hand-object interaction, and by using functional information derived from the object''s standard use. Children with ASD have no deficit in the second type of understanding, while they have difficulties in understanding others'' intentions when they have to rely exclusively on motor cues.     

Functional Brain Networks Develop from a “Local to Distributed” Organization

The mature human brain is organized into a collection of specialized functional networks that flexibly interact to support various cognitive functions. Studies of development often attempt to identify the organizing principles that guide the maturation of these functional networks. In this report, we combine resting state functional connectivity MRI (rs-fcMRI), graph analysis, community detection, and spring-embedding visualization techniques to analyze four separate networks defined in earlier studies. As we have previously reported, we find, across development, a trend toward ‘segregation’ (a general decrease in correlation strength) between regions close in anatomical space and ‘integration’ (an increased correlation strength) between selected regions distant in space. The generalization of these earlier trends across multiple networks suggests that this is a general developmental principle for changes in functional connectivity that would extend to large-scale graph theoretic analyses of large-scale brain networks. Communities in children are predominantly arranged by anatomical proximity, while communities in adults predominantly reflect functional relationships, as defined from adult fMRI studies. In sum, over development, the organization of multiple functional networks shifts from a local anatomical emphasis in children to a more “distributed” architecture in young adults. We argue that this “local to distributed” developmental characterization has important implications for understanding the development of neural systems underlying cognition. Further, graph metrics (e.g., clustering coefficients and average path lengths) are similar in child and adult graphs, with both showing “small-world”-like properties, while community detection by modularity optimization reveals stable communities within the graphs that are clearly different between young children and young adults. These observations suggest that early school age children and adults both have relatively efficient systems that may solve similar information processing problems in divergent ways.     

Functional Brain Networks Develop from a “Local to Distributed” Organization

The mature human brain is organized into a collection of specialized functional networks that flexibly interact to support various cognitive functions. Studies of development often attempt to identify the organizing principles that guide the maturation of these functional networks. In this report, we combine resting state functional connectivity MRI (rs-fcMRI), graph analysis, community detection, and spring-embedding visualization techniques to analyze four separate networks defined in earlier studies. As we have previously reported, we find, across development, a trend toward ‘segregation’ (a general decrease in correlation strength) between regions close in anatomical space and ‘integration’ (an increased correlation strength) between selected regions distant in space. The generalization of these earlier trends across multiple networks suggests that this is a general developmental principle for changes in functional connectivity that would extend to large-scale graph theoretic analyses of large-scale brain networks. Communities in children are predominantly arranged by anatomical proximity, while communities in adults predominantly reflect functional relationships, as defined from adult fMRI studies. In sum, over development, the organization of multiple functional networks shifts from a local anatomical emphasis in children to a more “distributed” architecture in young adults. We argue that this “local to distributed” developmental characterization has important implications for understanding the development of neural systems underlying cognition. Further, graph metrics (e.g., clustering coefficients and average path lengths) are similar in child and adult graphs, with both showing “small-world”-like properties, while community detection by modularity optimization reveals stable communities within the graphs that are clearly different between young children and young adults. These observations suggest that early school age children and adults both have relatively efficient systems that may solve similar information processing problems in divergent ways.     

Modeling spatial–temporal operations with context-dependent associative memories

We organize our behavior and store structured information with many procedures that require the coding of spatial and temporal order in specific neural modules. In the simplest cases, spatial and temporal relations are condensed in prepositions like “below” and “above”, “behind” and “in front of”, or “before” and “after”, etc. Neural operators lie beneath these words, sharing some similarities with logical gates that compute spatial and temporal asymmetric relations. We show how these operators can be modeled by means of neural matrix memories acting on Kronecker tensor products of vectors. The complexity of these memories is further enhanced by their ability to store episodes unfolding in space and time. How does the brain scale up from the raw plasticity of contingent episodic memories to the apparent stable connectivity of large neural networks? We clarify this transition by analyzing a model that flexibly codes episodic spatial and temporal structures into contextual markers capable of linking different memory modules.     

Neural basis of self and other representation in autism: an FMRI study of self-face recognition

Background

Autism is a developmental disorder characterized by decreased interest and engagement in social interactions and by enhanced self-focus. While previous theoretical approaches to understanding autism have emphasized social impairments and altered interpersonal interactions, there is a recent shift towards understanding the nature of the representation of the self in individuals with autism spectrum disorders (ASD). Still, the neural mechanisms subserving self-representations in ASD are relatively unexplored.

Methodology/Principal Findings

We used event-related fMRI to investigate brain responsiveness to images of the subjects'' own face and to faces of others. Children with ASD and typically developing (TD) children viewed randomly presented digital morphs between their own face and a gender-matched other face, and made “self/other” judgments. Both groups of children activated a right premotor/prefrontal system when identifying images containing a greater percentage of the self face. However, while TD children showed activation of this system during both self- and other-processing, children with ASD only recruited this system while viewing images containing mostly their own face.

Conclusions/Significance

This functional dissociation between the representation of self versus others points to a potential neural substrate for the characteristic self-focus and decreased social understanding exhibited by these individuals, and suggests that individuals with ASD lack the shared neural representations for self and others that TD children and adults possess and may use to understand others.     

Niche theory‐based modeling of assembly processes of viral communities in bats

Understanding the assembly processes of symbiont communities, including viromes and microbiomes, is important for improving predictions on symbionts’ biogeography and disease ecology. Here, we use phylogenetic, functional, and geographic filters to predict the similarity between symbiont communities, using as a test case the assembly process in viral communities of Mexican bats. We construct generalized linear models to predict viral community similarity, as measured by the Jaccard index, as a function of differences in host phylogeny, host functionality, and spatial co‐occurrence, evaluating the models using the Akaike information criterion. Two model classes are constructed: a “known” model, where virus–host relationships are based only on data reported in Mexico, and a “potential” model, where viral reports of all the Americas are used, but then applied only to bat species that are distributed in Mexico. Although the “known” model shows only weak dependence on any of the filters, the “potential” model highlights the importance of all three filter types—phylogeny, functional traits, and co‐occurrence—in the assemblage of viral communities. The differences between the “known” and “potential” models highlight the utility of modeling at different “scales” so as to compare and contrast known information at one scale to another one, where, for example, virus information associated with bats is much scarcer.     

Widespread Disruption of Functional Brain Organization in Early-Onset Alzheimer’s Disease

Early-onset Alzheimer’s disease (AD) patients present a different clinical profile than late-onset AD patients. This can be partially explained by cortical atrophy, although brain organization might provide more insight. The aim of this study was to examine functional connectivity in early-onset and late-onset AD patients. Resting-state fMRI scans of 20 early-onset (<65 years old), 28 late-onset (≥65 years old) AD patients and 15 “young” (<65 years old) and 31 “old” (≥65 years old) age-matched controls were available. Resting-state network-masks were used to create subject-specific maps. Group differences were examined using a non-parametric permutation test, accounting for gray-matter. Performance on five cognitive domains were used in a correlation analysis with functional connectivity in AD patients. Functional connectivity was not different in any of the RSNs when comparing the two control groups (young vs. old controls), which implies that there is no general effect of aging on functional connectivity. Functional connectivity in early-onset AD was lower in all networks compared to age-matched controls, where late-onset AD showed lower functional connectivity in the default-mode network. Functional connectivity was lower in early-onset compared to late-onset AD in auditory-, sensory-motor, dorsal-visual systems and the default mode network. Across patients, an association of functional connectivity of the default mode network was found with visuoconstruction. Functional connectivity of the right dorsal visual system was associated with attention across patients. In late-onset AD patients alone, higher functional connectivity of the sensory-motor system was associated with poorer memory performance. Functional brain organization was more widely disrupted in early-onset AD when compared to late-onset AD. This could possibly explain different clinical profiles, although more research into the relationship of functional connectivity and cognitive performance is needed.     

The Roles of FMRP-Regulated Genes in Autism Spectrum Disorder: Single- and Multiple-Hit Genetic Etiologies

Autism spectrum disorder (ASD) is a highly heritable complex neurodevelopmental condition characterized by impairments in social interaction and communication and restricted and repetitive behaviors. Although roles for both de novo and familial genetic variation have been documented, the underlying disease mechanisms remain poorly elucidated. In this study, we defined and explored distinct etiologies of genetic variants that affect genes regulated by Fragile-X mental retardation protein (FMRP), thought to play a key role in neuroplasticity and neuronal translation, in ASD-affected individuals. In particular, we developed the Trend test, a pathway-association test that is able to robustly detect multiple-hit etiologies and is more powerful than existing approaches. Exploiting detailed spatiotemporal maps of gene expression within the human brain, we identified four discrete FMRP-target subpopulations that exhibit distinct functional biases and contribute to ASD via different types of genetic variation. We also demonstrated that FMRP target genes are more likely than other genes with similar expression patterns to contribute to disease. We developed the hypothesis that FMRP targets contribute to ASD via two distinct etiologies: (1) ultra-rare and highly penetrant single disruptions of embryonically upregulated FMRP targets (“single-hit etiology”) or (2) the combination of multiple less penetrant disruptions of nonembryonic, synaptic FMRP targets (“multiple-hit etiology”). The Trend test provides rigorous support for a multiple-hit genetic etiology in a subset of autism cases and is easily extendible to combining information from multiple types of genetic variation (i.e., copy-number and exome variants), increasing its value to next-generation sequencing approaches.     

Network analysis of intrinsic functional brain connectivity in Alzheimer's disease

Functional brain networks detected in task-free (“resting-state”) functional magnetic resonance imaging (fMRI) have a small-world architecture that reflects a robust functional organization of the brain. Here, we examined whether this functional organization is disrupted in Alzheimer's disease (AD). Task-free fMRI data from 21 AD subjects and 18 age-matched controls were obtained. Wavelet analysis was applied to the fMRI data to compute frequency-dependent correlation matrices. Correlation matrices were thresholded to create 90-node undirected-graphs of functional brain networks. Small-world metrics (characteristic path length and clustering coefficient) were computed using graph analytical methods. In the low frequency interval 0.01 to 0.05 Hz, functional brain networks in controls showed small-world organization of brain activity, characterized by a high clustering coefficient and a low characteristic path length. In contrast, functional brain networks in AD showed loss of small-world properties, characterized by a significantly lower clustering coefficient (p<0.01), indicative of disrupted local connectivity. Clustering coefficients for the left and right hippocampus were significantly lower (p<0.01) in the AD group compared to the control group. Furthermore, the clustering coefficient distinguished AD participants from the controls with a sensitivity of 72% and specificity of 78%. Our study provides new evidence that there is disrupted organization of functional brain networks in AD. Small-world metrics can characterize the functional organization of the brain in AD, and our findings further suggest that these network measures may be useful as an imaging-based biomarker to distinguish AD from healthy aging.     

Explicating the Face Perception Network with White Matter Connectivity

A network of multiple brain regions is recruited in face perception. Our understanding of the functional properties of this network can be facilitated by explicating the structural white matter connections that exist between its functional nodes. We accomplished this using functional MRI (fMRI) in combination with fiber tractography on high angular resolution diffusion weighted imaging data. We identified the three nodes of the core face network: the “occipital face area” (OFA), the “fusiform face area” (mid-fusiform gyrus or mFus), and the superior temporal sulcus (STS). Additionally, a region of the anterior temporal lobe (aIT), implicated as being important for face perception was identified. Our data suggest that we can further divide the OFA into multiple anatomically distinct clusters – a partitioning consistent with several recent neuroimaging results. More generally, structural white matter connectivity within this network revealed: 1) Connectivity between aIT and mFus, and between aIT and occipital regions, consistent with studies implicating this posterior to anterior pathway as critical to normal face processing; 2) Strong connectivity between mFus and each of the occipital face-selective regions, suggesting that these three areas may subserve different functional roles; 3) Almost no connectivity between STS and mFus, or between STS and the other face-selective regions. Overall, our findings suggest a re-evaluation of the “core” face network with respect to what functional areas are or are not included in this network.     

Integrated Analysis and Visualization of Group Differences in Structural and Functional Brain Connectivity: Applications in Typical Ageing and Schizophrenia

Structural and functional brain connectivity are increasingly used to identify and analyze group differences in studies of brain disease. This study presents methods to analyze uni- and bi-modal brain connectivity and evaluate their ability to identify differences. Novel visualizations of significantly different connections comparing multiple metrics are presented. On the global level, “bi-modal comparison plots” show the distribution of uni- and bi-modal group differences and the relationship between structure and function. Differences between brain lobes are visualized using “worm plots”. Group differences in connections are examined with an existing visualization, the “connectogram”. These visualizations were evaluated in two proof-of-concept studies: (1) middle-aged versus elderly subjects; and (2) patients with schizophrenia versus controls. Each included two measures derived from diffusion weighted images and two from functional magnetic resonance images. The structural measures were minimum cost path between two anatomical regions according to the “Statistical Analysis of Minimum cost path based Structural Connectivity” method and the average fractional anisotropy along the fiber. The functional measures were Pearson’s correlation and partial correlation of mean regional time series. The relationship between structure and function was similar in both studies. Uni-modal group differences varied greatly between connectivity types. Group differences were identified in both studies globally, within brain lobes and between regions. In the aging study, minimum cost path was highly effective in identifying group differences on all levels; fractional anisotropy and mean correlation showed smaller differences on the brain lobe and regional levels. In the schizophrenia study, minimum cost path and fractional anisotropy showed differences on the global level and within brain lobes; mean correlation showed small differences on the lobe level. Only fractional anisotropy and mean correlation showed regional differences. The presented visualizations were helpful in comparing and evaluating connectivity measures on multiple levels in both studies.     

Dynamically Allocated Hub in Task-Evoked Network Predicts the Vulnerable Prefrontal Locus for Contextual Memory Retrieval in Macaques

Neuroimaging and neurophysiology have revealed that multiple areas in the prefrontal cortex (PFC) are activated in a specific memory task, but severity of impairment after PFC lesions is largely different depending on which activated area is damaged. The critical relationship between lesion sites and impairments has not yet been given a clear mechanistic explanation. Although recent works proposed that a whole-brain network contains hubs that play integrative roles in cortical information processing, this framework relying on an anatomy-based structural network cannot account for the vulnerable locus for a specific task, lesioning of which would bring impairment. Here, we hypothesized that (i) activated PFC areas dynamically form an ordered network centered at a task-specific “functional hub” and (ii) the lesion-effective site corresponds to the “functional hub,” but not to a task-invariant “structural hub.” To test these hypotheses, we conducted functional magnetic resonance imaging experiments in macaques performing a temporal contextual memory task. We found that the activated areas formed a hierarchical hub-centric network based on task-evoked directed connectivity, differently from the anatomical network reflecting axonal projection patterns. Using a novel simulated-lesion method based on support vector machine, we estimated severity of impairment after lesioning of each area, which accorded well with a known dissociation in contextual memory impairment in macaques (impairment after lesioning in area 9/46d, but not in area 8Ad). The predicted severity of impairment was proportional to the network “hubness” of the virtually lesioned area in the task-evoked directed connectivity network, rather than in the anatomical network known from tracer studies. Our results suggest that PFC areas dynamically and cooperatively shape a functional hub-centric network to reallocate the lesion-effective site depending on the cognitive processes, apart from static anatomical hubs. These findings will be a foundation for precise prediction of behavioral impacts of damage or surgical intervention in human brains.     

Long-Term Effects of Temporal Lobe Epilepsy on Local Neural Networks: A Graph Theoretical Analysis of Corticography Recordings

Purpose

Pharmaco-resistant temporal lobe epilepsy (TLE) is often treated with surgical intervention at some point. As epilepsy surgery is considered a last resort by most physicians, a long history of epileptic seizures prior to surgery is not uncommon. Little is known about the effects of ongoing TLE on neural functioning. A better understanding of these effects might influence the moment of surgical intervention. Functional connectivity (interaction between spatially distributed brain areas) and network structure (integration and segregation of information processing) are thought to be essential for optimal brain functioning. We report on the impact of TLE duration on temporal lobe functional connectivity and network characteristics.

Methods

Functional connectivity of the temporal lobe at the time of surgery was assessed by means of interictal electrocorticography (ECoG) recordings of 27 TLE patients by using the phase lag index (PLI). Graphs (abstract network representations) were reconstructed from the PLI matrix and characterized by the clustering coefficient C (local clustering), the path length L (overall network interconnectedness), and the “small world index” S (network configuration).

Results

Functional connectivity (average PLI), clustering coefficients, and the small world index were negatively correlated with TLE duration in the broad frequency band (0.5–48 Hz).

Discussion

Temporal lobe functional connectivity is lower in patients with longer TLE history, and longer TLE duration is correlated with more random network configuration. Our findings suggest that the neural networks of TLE patients become more pathological over time, possibly due to temporal lobe changes associated with long-standing lesional epilepsy.     

Influence of the Time of Day and Fasting Duration on Glucose Level following a 1-Hour, 50-Gram Glucose Challenge Test in Pregnant Women

Background

Previous studies have shown that the time of day (TD) of glucose measurement and the fasting duration (FD) influence the glucose levels in adults. Few studies have examined the effects of the TD and FD on the glucose level following a 1-hour, 50-gram glucose challenge test (GCT) in pregnant women in screening for or diagnosing gestational diabetes mellitus (GDM). The objective of this study was to investigate the influence of the TD (morning, afternoon, night) and the FD (the time of the last food ingestion as follows: ≤1 hour, 1–2 hours, and >2 hours) by examining their combined effects on the glucose levels following a 50-gram GCT in pregnant women.

Methods and Results

We analyzed the data of 1,454 non-diabetic pregnant Taiwanese women in a prospective study. Multiple linear regression and multiple logistic regression were used to estimate the relationships between the 9 TD-FD groups and the continuous and binary glucose levels (cut-off at 140 mg/dL) following a 50-gram GCT, after adjusting for maternal age, nulliparity, pre-pregnancy body mass index, and weight gain. Different TD and FD groups were associated with variable glucose responses to the 50-gram GCT, some of which were significant. The estimate coefficients (β) of the TD-FD groups “night, ≤1 hr” and “night, 1–2 hr” revealed significantly lower glucose concentrations [β (95% confidence interval [CI]): −6.46 (−12.53, −0.38) and −6.85 (−12.50, −1.20)] compared with the “morning, >2 hr” group. The TD-FD groups “afternoon, ≤1 hr” and “afternoon, 1–2 hr” showed significantly lower odds ratios (OR) of a positive GCT; the adjusted ORs (95% CI) were 0.54 (0.31–0.95) and 0.58 (0.35–0.96), respectively.

Conclusions

Our findings demonstrate the importance of standardizing the TD and FD for the 1-hour, 50-gram GCT. In screening for and diagnosing GDM, the TD and FD are modifiable factors that should be considered in clinical practice and epidemiological studies.     

Intrinsic connectivity reveals functionally distinct cortico-hippocampal networks in the human brain

Episodic memory depends on interactions between the hippocampus and interconnected neocortical regions. Here, using data-driven analyses of resting-state functional magnetic resonance imaging (fMRI) data, we identified the networks that interact with the hippocampus—the default mode network (DMN) and a “medial temporal network” (MTN) that included regions in the medial temporal lobe (MTL) and precuneus. We observed that the MTN plays a critical role in connecting the visual network to the DMN and hippocampus. The DMN could be further divided into 3 subnetworks: a “posterior medial” (PM) subnetwork comprised of posterior cingulate and lateral parietal cortices; an “anterior temporal” (AT) subnetwork comprised of regions in the temporopolar and dorsomedial prefrontal cortex; and a “medial prefrontal” (MP) subnetwork comprised of regions primarily in the medial prefrontal cortex (mPFC). These networks vary in their functional connectivity (FC) along the hippocampal long axis and represent different kinds of information during memory-guided decision-making. Finally, a Neurosynth meta-analysis of fMRI studies suggests new hypotheses regarding the functions of the MTN and DMN subnetworks, providing a framework to guide future research on the neural architecture of episodic memory.

Episodic memory depends on interactions between the hippocampus and interconnected neocortical regions. This study uses network analyses of intrinsic brain networks at rest to identify and characterize brain networks that interact with the hippocampus and have distinct functions during memory-guided decision making.     

Cognitive Demands of Lower Paleolithic Toolmaking

Stone tools provide some of the most abundant, continuous, and high resolution evidence of behavioral change over human evolution, but their implications for cognitive evolution have remained unclear. We investigated the neurophysiological demands of stone toolmaking by training modern subjects in known Paleolithic methods (“Oldowan”, “Acheulean”) and collecting structural and functional brain imaging data as they made technical judgments (outcome prediction, strategic appropriateness) about planned actions on partially completed tools. Results show that this task affected neural activity and functional connectivity in dorsal prefrontal cortex, that effect magnitude correlated with the frequency of correct strategic judgments, and that the frequency of correct strategic judgments was predictive of success in Acheulean, but not Oldowan, toolmaking. This corroborates hypothesized cognitive control demands of Acheulean toolmaking, specifically including information monitoring and manipulation functions attributed to the "central executive" of working memory. More broadly, it develops empirical methods for assessing the differential cognitive demands of Paleolithic technologies, and expands the scope of evolutionary hypotheses that can be tested using the available archaeological record.     

A Bayesian Model of the Uncanny Valley Effect for Explaining the Effects of Therapeutic Robots in Autism Spectrum Disorder

One of the core features of autism spectrum disorder (ASD) is impaired reciprocal social interaction, especially in processing emotional information. Social robots are used to encourage children with ASD to take the initiative and to interact with the robotic tools to stimulate emotional responses. However, the existing evidence is limited by poor trial designs. The purpose of this study was to provide computational evidence in support of robot-assisted therapy for children with ASD. We thus propose an emotional model of ASD that adapts a Bayesian model of the uncanny valley effect, which holds that a human-looking robot can provoke repulsion and sensations of eeriness. Based on the unique emotional responses of children with ASD to the robots, we postulate that ASD induces a unique emotional response curve, more like a cliff than a valley. Thus, we performed numerical simulations of robot-assisted therapy to evaluate its effects. The results showed that, although a stimulus fell into the uncanny valley in the typical condition, it was effective at avoiding the uncanny cliff in the ASD condition. Consequently, individuals with ASD may find it more comfortable, and may modify their emotional response, if the robots look like deformed humans, even if they appear “creepy” to typical individuals. Therefore, we suggest that our model explains the effects of robot-assisted therapy in children with ASD and that human-looking robots may have potential advantages for improving social interactions in ASD.     

Changes in Task-Related Functional Connectivity across Multiple Spatial Scales Are Related to Reading Performance

Reading requires the interaction of a distributed set of cortical areas whose distinct patterns give rise to a wide range of individual skill. However, the nature of these neural interactions and their relation to reading performance are still poorly understood. Functional connectivity analyses of fMRI data can be used to characterize the nature of interactivity of distributed brain networks, yet most previous studies have focused on connectivity during task-free (i.e., “resting state”) conditions. Here, we report new methods for assessing task-related functional connectivity using data-driven graph theoretical methods and describe how large-scale patterns of connectivity relate to individual variability in reading performance among children. We found that connectivity patterns of subjects performing a reading task could be decomposed hierarchically into multiple sub-networks, and we observed stronger long-range interaction between sub-networks in subjects with higher task accuracy. Additionally, we found a network of hub regions known to be critical to reading that displays increased short-range synchronization in higher accuracy subjects. These individual differences in task-related functional connectivity reveal that increased interaction between distant regions, coupled with selective local integration within key regions, is associated with better reading performance. Importantly, we show that task-related neuroimaging data contains far more information than usually extracted via standard univariate analyses – information that can meaningfully relate neural connectivity patterns to cognition and task.     

The gradient clusteron: A model neuron that learns to solve classification tasks via dendritic nonlinearities,structural plasticity,and gradient descent

Synaptic clustering on neuronal dendrites has been hypothesized to play an important role in implementing pattern recognition. Neighboring synapses on a dendritic branch can interact in a synergistic, cooperative manner via nonlinear voltage-dependent mechanisms, such as NMDA receptors. Inspired by the NMDA receptor, the single-branch clusteron learning algorithm takes advantage of location-dependent multiplicative nonlinearities to solve classification tasks by randomly shuffling the locations of “under-performing” synapses on a model dendrite during learning (“structural plasticity”), eventually resulting in synapses with correlated activity being placed next to each other on the dendrite. We propose an alternative model, the gradient clusteron, or G-clusteron, which uses an analytically-derived gradient descent rule where synapses are "attracted to" or "repelled from" each other in an input- and location-dependent manner. We demonstrate the classification ability of this algorithm by testing it on the MNIST handwritten digit dataset and show that, when using a softmax activation function, the accuracy of the G-clusteron on the all-versus-all MNIST task (~85%) approaches that of logistic regression (~93%). In addition to the location update rule, we also derive a learning rule for the synaptic weights of the G-clusteron (“functional plasticity”) and show that a G-clusteron that utilizes the weight update rule can achieve ~89% accuracy on the MNIST task. We also show that a G-clusteron with both the weight and location update rules can learn to solve the XOR problem from arbitrary initial conditions.