CBGTPy: An extensible cortico-basal ganglia-thalamic framework for modeling biological decision making

Abstract

Here we introduce CBGTPy, a virtual environment for designing and testing goal-directed agents with internal dynamics that are modeled off of the cortico-basal-ganglia-thalamic (CBGT) pathways in the mammalian brain. CBGTPy enables researchers to investigate the internal dynamics of the CBGT system during a variety of tasks, allowing for the formation of testable predictions about animal behavior and neural activity. The framework has been designed around the principle of flexibility, such that many experimental parameters in a decision making paradigm can be easily defined and modified. Here we demonstrate the capabilities of CBGTPy across a range of single and multi-choice tasks, highlighting the ease of set up and the biologically realistic behavior that it produces. We show that CBGTPy is extensible enough to apply to a wide range of experimental protocols and to allow for the implementation of model extensions with minimal developmental effort.

Distinct dopaminergic spike-timing-dependent plasticity rules are suited to different functional roles

“…We now revisit the question of whether to use a single eligibility trace summing up both positive (corresponding to pre-before-post spike pairs) and negative (post-before-pre) contributions, as is done in [31], or to use two different traces for the positive and negative components, as we do above. We test a single-trace version of our model that replaces Eq (3) with where γ ≥ 1 is a scaling parameter controlling the strength of negative eligibility terms relative to positive terms.…”

“…We test a single-trace version of our model that replaces Eq (3) with where γ ≥ 1 is a scaling parameter controlling the strength of negative eligibility terms relative to positive terms. The single-trace differential equation for the weights in the additive and multiplicative cases is given by and for the corticostriatal model is given by The single-trace version of the corticostriatal model is largely equivalent to the model in described in [31], although they use different scaling factors and time constants for pre- and postsynaptic activity.…”

“…The additive and multiplicative models are based on models of STDP (without dopamine) described in [25,26] (for the additive model) and [27][28][29] (for the multiplicative model); we mainly follow the presentation in [30]. The corticostriatal model is based on a computational model of the corticostriatal connections described in [31], specifically connections onto striatal spiny projection neurons that express the D1 dopamine receptor, sometimes referred to as direct pathway SPNs.…”

Arkypallidal neurons in the external globus pallidus can mediate inhibitory control by altering competition in the striatum

“…This third pathway is known to operate independently of the striatum, directly regulating thalamic inhibition through cortical input to the STN and its excitatory influence on GPi. For more details on the network implementation, see [38]; see also [11] for a review of experimental findings that justify the network structure used.…”

“…This current amplifies the activity of the targeted region. For a comprehensive understanding of the task, refer to our methods paper [38]. The stop signal current is characterized by several key parameters: (a) amplitude , indicating the intensity of the applied stimulation; (b) population , specifying the targeted CBGT region; (c) onset , indicating the timing of the stimulation relative to the trial onset; (d) duration , defining how long the stimulation lasts.…”