Synaptic dynamics differ markedly across connections and strongly regulate how action potentials are being communicated. To model the range of synaptic dynamics observed in experiments, we develop a flexible mathematical framework based on a linear-nonlinear operation. This model can capture various experimentally observed features of synaptic dynamics and different types of heteroskedasticity. Despite its conceptual simplicity, we show it is more adaptable than previous models. Combined with a standard maximum likelihood approach, synaptic dynamics can be accurately and efficiently characterized using naturalistic stimulation patterns. These results make explicit that synaptic processing bears algorithmic similarities with information processing in convolutional neural networks. Author summaryUnderstanding how information is transmitted relies heavily on knowledge of the underlying regulatory synaptic dynamics. Existing computational models for capturing such dynamics are often either very complex or too restrictive. As a result, effectively capturing the different types of dynamics observed experimentally remains a challenging problem. Here, we propose a mathematically flexible linear-nonlinear model that is capable of efficiently characterizing synaptic dynamics. We demonstrate the ability of this model to capture different features of experimentally observed data.The nervous system has evolved a communication system largely based on temporal 2 sequences of action potentials. A central feature of this communication is that action 3 potentials are communicated with variable efficacy on short (10 ms -10 s) time 4 scales [1-6]. The dynamics of synaptic efficacy at short time scales, or short-term 5 plasticity (STP), can be a powerful determinant of the flow of information, allowing the 6 same axon to communicate independent messages to different post-synaptic 7 targets [7, 8]. Properties of STP vary markedly across projections [9-11], leading to the 8idea that connections can be conceived as belonging to distinct classes [12,13] and that 9 these distinct classes shape information transmission in vivo [14][15][16]. Thus, to 10 May 29, 2020 1/25 understand the flow of information in neuronal networks, the connectome must be 11 indexed with an accurate description of STP properties. 12One approach to characterizing synaptic dynamics is to perform targeted 13 experiments and extract a summary feature, most commonly the paired-pulse 14 ratio [5,[17][18][19], whereby a synapse can be classified as short-term depressing (STD) or 15 short-term facilitating (STF). However, a single summary feature is insufficient to 16 capture the full extent of STP diversity. Longer or more complex stimulation patterns 17 are required to describe delayed facilitation onset [6], biphasic STP [20, 21] or the 18 distinction between supra-and sub-linear facilitation [22]. Such atypical STP dynamics 19 challenge the traditional dichotomy of STF and STD and suggest that more complex 20 phenotypes can exist and contribute to network function in...
Short-term synaptic dynamics differ markedly across connections and strongly regulate how action potentials communicate information. To model the range of synaptic dynamics observed in experiments, we have developed a flexible mathematical framework based on a linear-nonlinear operation. This model can capture various experimentally observed features of synaptic dynamics and different types of heteroskedasticity. Despite its conceptual simplicity, we show that it is more adaptable than previous models. Combined with a standard maximum likelihood approach, synaptic dynamics can be accurately and efficiently characterized using naturalistic stimulation patterns. These results make explicit that synaptic processing bears algorithmic similarities with information processing in convolutional neural networks.
Synapses show preferential responses to particular temporal patterns of activity. Across individual synapses, there is a large degree of response heterogeneity that is informally or tacitly separated into classes, and typically only two: facilitating and depressing short-term plasticity. Here we combined a kernel-based model and machine learning techniques to infer the number and the characteristics of functionally distinct subtypes of short-term synaptic dynamics in a large dataset of glutamatergic cortical connections. To this end, we took two independent approaches. First, we used unsupervised techniques to group similar synapses into clusters. Second, we used supervised prediction of cell subclasses to reveal features of synaptic dynamics that characterized cellular genetic profiles. In rodent data, we found five clusters with a remarkable degree of convergence with the transgenic-associated subtypes. Two of these clusters corresponded to different degrees of facilitation, two corresponded to depression with different degrees of variability and one corresponded to depression-then-facilitation. Strikingly, the application of the same clustering method in human data inferred highly similar clusters to those observed in rodents, supportive of a stable clustering procedure and suggesting a homology of functional subtypes across species. This nuanced dictionary of functional subtypes shapes the heterogeneity of cortical synaptic dynamics and provides a lens into the basic motifs of information transmission in the brain.
In various scientific fields, researchers make use of partitioning methods (e.g., K-means) to disclose the structural mechanisms underlying object by variable data. In some instances, however, a grouping of objects into clusters that are allowed to overlap (i.e., assigning objects to multiple clusters) might lead to a better representation of the underlying clustering structure. To obtain an overlapping object clustering from object by variable data, Mirkin’s ADditive PROfile CLUStering (ADPROCLUS) model may be used. A major challenge when performing ADPROCLUS is to determine the optimal number of overlapping clusters underlying the data, which pertains to a model selection problem. Up to now, however, this problem has not been systematically investigated and almost no guidelines can be found in the literature regarding appropriate model selection strategies for ADPROCLUS. Therefore, in this paper, several existing model selection strategies for K-means (a.o., CHull, the Caliński-Harabasz, Krzanowski-Lai, Average Silhouette Width and Dunn Index and information-theoretic measures like AIC and BIC) and two cross-validation based strategies are tailored towards an ADPROCLUS context and are compared to each other in an extensive simulation study. The results demonstrate that CHull outperforms all other model selection strategies and this especially when the negative log-likelihood, which is associated with a minimal stochastic extension of ADPROCLUS, is used as (mis)fit measure. The analysis of a post hoc AIC-based model selection strategy revealed that better performance may be obtained when a different—more appropriate—definition of model complexity for ADPROCLUS is used.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.