Inferring circuit mechanisms from sparse neural recording and global perturbation in grid cells
- The University of California, United States
- The University of Texas, United States
Figures
Figure 1 with 2 supplements
Mechanistically distinct models that cannot be ruled out with existing results.
(a–d) Recurrent pattern-forming models. Gray bumps: population activity profiles. Blue: Profile of synaptic weights from a representative grid cell (green) to the rest of the network. Bottom of each …
Figure 1—figure supplement 1
(Weakly) coupling neurons based on periodic activity patterning converts an aperiodic network into an effectively fully periodic one.
(a) The population pattern period in an aperiodic network expands continuously with increasing inhibition strength () over a range. The ordinate shows the stretch-factor , which quantifies the …
Figure 1—figure supplement 2
The a priori theoretical implausibility of partially periodic networks.
Population activity in the cortical sheet (yellow-black blobs), with schematic of connectivity (green). Note that in the bulk of the sheet, connectivity is local and not determined by the periodic …
Figure 2 with 4 supplements
Global perturbation with analysis of phase shifts: signatures of recurrent patterning.
(a) Schematic of population activity before (blue) and after (red) a 10% period expansion (; the center of expansion is shown at left, but results are independent of this choice) in an aperiodic …
Figure 2—figure supplement 1
Dynamical simulations of the aperiodic network with LNP dynamics: gradual change in population period.
Change in population pattern period as the the time-constant (a) and inhibition strength (b) are increased in a 1D aperiodic network (see Materials and methods). In all trials (black circles), the …
Figure 2—figure supplement 2
When the 2:1 relationship between number of peaks in the DRPS and the number of bumps in the population pattern breaks down.
Top: Schematic of the phase in a population pattern, pre- (blue) and post- (red) perturbation, for a large 1D network with many bumps with population period stretch factor =0.1, with phase shift …
Figure 2—figure supplement 3
Alternative formulation of the DRPS.
(a) 1D population activity, pre- (blue) and post-perturbation, for a increase the wavelength of the pattern ( neurons), with pattern expansion is centered at the left network edge. Circle, …
Figure 2—figure supplement 4
Cortical Hodgkin-Huxley (CHH) simulations to assess the effects of cooling as an experimental perturbation and to elucidate the link between temperature and parameter settings in grid cell models with simpler neurons.
Results from a 1D aperiodic recurrent network grid cell model with CHH neurons. In all simulations, the network is driven by a constant velocity of 0.3 m/s and the run is a 10 s trajectory. (a–d) …
Figure 3 with 1 supplement
Effects of perturbation in recurrent and feedforward neural network simulations: predictions for experiment.
(a) Simulations of aperiodic (column 1), partially periodic (column 2), and fully periodic (column 3) networks show changes in the population pattern pre-perturbation (first row; ) to …
Figure 3—figure supplement 1
Changes in spatial tuning period in neural network simulations of the grid cell circuit are due to changes in both the population period and the velocity response of the network.
Change in spatial tuning period (a), population pattern period (b), and the velocity response (c) for the different network architectures (see Materials and methods for definitions of measures). …
Figure 4 with 1 supplement
Data limitations and the resolvability of predictions.
(a) Left: The quantal structure of the DRPS (along first principal axis of the 2D phase) is apparent even in small samples of the population (black: full population; red: n = 10 cells out of 1600; …
Figure 4—figure supplement 1
Effects of uncertainty in phase estimation.
(a) Copied from Figure 4b. First and second columns: DRPS (200 bins; gray line: raw; black line: smoothed with 2-bin Gaussian) for different numbers of population pattern bumps along the first …
Figure 5 with 5 supplements
Decision tree for experimentally discriminating circuit mechanisms.
The ‘specific’ approach involves a specific perturbation to either the gain of inhibition or the neural time-constants. Under the assumption of this kind of perturbation, the period, the amplitude, …
Figure 5—figure supplement 1
Perturbations applied to a random subset of neurons in the network.
(a) Population period as a function of a fractional perturbation of the network for aperiodic (top row), partially periodic (middle row), and fully periodic (bottom row) networks. Black circles …
Figure 5—figure supplement 2
Perturbations applied separately to the excitatory and inhibitory populations.
(a) Left: Population period as a function of a global perturbation of the synaptic time constants (, where ms and is the perturbation parameter scale factor), for aperiodic (top row), …
Figure 5—figure supplement 3
Perturbations applied to the gain of the neural response.
(a) Left: Population period as a function of a global perturbation of the firing rates (the perturbation scale factor is applied multiplicatively to both and , see Materials and methods), …
Figure 5—figure supplement 4
The effects of perturbations in networks with spatially untuned inhibitory neurons.
(a) Left: Snapshot of the I (black), E (red), and E (blue) population activities, for the case when EE connections are added (i.e., E-E, E-E, E-E, E-E - see Materials and methods …
Figure 5—figure supplement 5
Learned place cell-based intrinsic error correction/resetting of grid phase in familiar environments is not predicted to play an important role in novel environments according to model.
(a) Top row: Firing field of a single place cell (cell 67) learned in two familiar environments (first and second column) based on associating this field with the co-active grid cells (see Materials …
Additional files
-
Transparent reporting form
- https://doi.org/10.7554/eLife.33503.020
