Limits...
Low-noise encoding of active touch by layer 4 in the somatosensory cortex.

Hires SA, Gutnisky DA, Yu J, O'Connor DH, Svoboda K - Elife (2015)

Bottom Line: The amount of irreducible internal noise in spike trains, an important constraint on models of cortical networks, has been difficult to estimate, since behavior and brain state must be precisely controlled or tracked.The variance of touch responses was smaller than expected from Poisson processes, often reaching the theoretical minimum.Layer 4 spike trains thus reflect the millisecond-timescale structure of tactile input with little noise.

View Article: PubMed Central - PubMed

Affiliation: Janelia Research Campus, Howard Hughes Medical Institute, Ashburn, United States.

ABSTRACT
Cortical spike trains often appear noisy, with the timing and number of spikes varying across repetitions of stimuli. Spiking variability can arise from internal (behavioral state, unreliable neurons, or chaotic dynamics in neural circuits) and external (uncontrolled behavior or sensory stimuli) sources. The amount of irreducible internal noise in spike trains, an important constraint on models of cortical networks, has been difficult to estimate, since behavior and brain state must be precisely controlled or tracked. We recorded from excitatory barrel cortex neurons in layer 4 during active behavior, where mice control tactile input through learned whisker movements. Touch was the dominant sensorimotor feature, with >70% spikes occurring in millisecond timescale epochs after touch onset. The variance of touch responses was smaller than expected from Poisson processes, often reaching the theoretical minimum. Layer 4 spike trains thus reflect the millisecond-timescale structure of tactile input with little noise.

No MeSH data available.


Related in: MedlinePlus

Grouping of touch events using density-based clustering (OPTICS algorithm; see ‘Materials and methods’) for an example neuron.(A) Separation into five groups of touches by successively removing touches with similar attributes. In the first panel, blue dots represents the 20% of touches with similar velocity at touch and maximum curvature change (maximum between 0–20 ms after touch). Black dots represent all the other touches. The second group of touches (red) is obtained by repeating the clustering with the blue dots removed. This process is repeated until obtaining the five touch groups. (first group = blue; second group = red; third group = green; fourth group = cyan; fifth group = magenta). (B) Standard deviation in each of the five groups compared to random sampling. The procedure finds subset of points with significantly less behavioral variability than in the full dataset. The black curve represents the standard deviation for velocity at touch (z-scored), the curve for maximum curvature is similar (not shown).DOI:http://dx.doi.org/10.7554/eLife.06619.016
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4525079&req=5

fig6s1: Grouping of touch events using density-based clustering (OPTICS algorithm; see ‘Materials and methods’) for an example neuron.(A) Separation into five groups of touches by successively removing touches with similar attributes. In the first panel, blue dots represents the 20% of touches with similar velocity at touch and maximum curvature change (maximum between 0–20 ms after touch). Black dots represent all the other touches. The second group of touches (red) is obtained by repeating the clustering with the blue dots removed. This process is repeated until obtaining the five touch groups. (first group = blue; second group = red; third group = green; fourth group = cyan; fifth group = magenta). (B) Standard deviation in each of the five groups compared to random sampling. The procedure finds subset of points with significantly less behavioral variability than in the full dataset. The black curve represents the standard deviation for velocity at touch (z-scored), the curve for maximum curvature is similar (not shown).DOI:http://dx.doi.org/10.7554/eLife.06619.016

Mentions: To sort out external and intrinsic contributions to spike count variability, we took several steps to reduce external variability due to differences in sensorimotor variables. Touch responses in L4 neurons are modulated by adaptation, pretouch velocity and curvature of the whisker (which is proportional to touch force) (Figure 5). To reduce adaptation effects we selected touch epochs in which the inter contact interval (ICI) was longer than 250 ms. We divided the remaining touch events into five groups (N, number of points per group; minimum, 20), clustered by touch characteristics. We z-scored pretouch velocity and the maximum curvature change shortly after touch (0–20 ms). For each neuron we clustered the touch events using a density-based clustering method (OPTICS; [Cunningham and Yu, 2014]). The output of OPTICS gives an ordered list of points sorted by similarity. We searched for the set of consecutive sorted points that mimimized the sum over all pairwise distances. After obtaining the set of touch events for the first bin, we removed those points and proceeded in the same manner to obtain the second data bin. We repeated the procedure until obtaining five data bins (Figure 6—figure supplement 1). We calculated the FF by counting spikes in sliding windows of 10 ms for each cell and each of the five bins (Figure 6). Confidence intervals for the FFs were obtained by resampling 1000 times. We also computed FFs using a Poisson process with absolute refractory period, with average touch response matched to the data (Berry II and Meister, 1998) (Figure 2—figure supplement 1).


Low-noise encoding of active touch by layer 4 in the somatosensory cortex.

Hires SA, Gutnisky DA, Yu J, O'Connor DH, Svoboda K - Elife (2015)

Grouping of touch events using density-based clustering (OPTICS algorithm; see ‘Materials and methods’) for an example neuron.(A) Separation into five groups of touches by successively removing touches with similar attributes. In the first panel, blue dots represents the 20% of touches with similar velocity at touch and maximum curvature change (maximum between 0–20 ms after touch). Black dots represent all the other touches. The second group of touches (red) is obtained by repeating the clustering with the blue dots removed. This process is repeated until obtaining the five touch groups. (first group = blue; second group = red; third group = green; fourth group = cyan; fifth group = magenta). (B) Standard deviation in each of the five groups compared to random sampling. The procedure finds subset of points with significantly less behavioral variability than in the full dataset. The black curve represents the standard deviation for velocity at touch (z-scored), the curve for maximum curvature is similar (not shown).DOI:http://dx.doi.org/10.7554/eLife.06619.016
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4525079&req=5

fig6s1: Grouping of touch events using density-based clustering (OPTICS algorithm; see ‘Materials and methods’) for an example neuron.(A) Separation into five groups of touches by successively removing touches with similar attributes. In the first panel, blue dots represents the 20% of touches with similar velocity at touch and maximum curvature change (maximum between 0–20 ms after touch). Black dots represent all the other touches. The second group of touches (red) is obtained by repeating the clustering with the blue dots removed. This process is repeated until obtaining the five touch groups. (first group = blue; second group = red; third group = green; fourth group = cyan; fifth group = magenta). (B) Standard deviation in each of the five groups compared to random sampling. The procedure finds subset of points with significantly less behavioral variability than in the full dataset. The black curve represents the standard deviation for velocity at touch (z-scored), the curve for maximum curvature is similar (not shown).DOI:http://dx.doi.org/10.7554/eLife.06619.016
Mentions: To sort out external and intrinsic contributions to spike count variability, we took several steps to reduce external variability due to differences in sensorimotor variables. Touch responses in L4 neurons are modulated by adaptation, pretouch velocity and curvature of the whisker (which is proportional to touch force) (Figure 5). To reduce adaptation effects we selected touch epochs in which the inter contact interval (ICI) was longer than 250 ms. We divided the remaining touch events into five groups (N, number of points per group; minimum, 20), clustered by touch characteristics. We z-scored pretouch velocity and the maximum curvature change shortly after touch (0–20 ms). For each neuron we clustered the touch events using a density-based clustering method (OPTICS; [Cunningham and Yu, 2014]). The output of OPTICS gives an ordered list of points sorted by similarity. We searched for the set of consecutive sorted points that mimimized the sum over all pairwise distances. After obtaining the set of touch events for the first bin, we removed those points and proceeded in the same manner to obtain the second data bin. We repeated the procedure until obtaining five data bins (Figure 6—figure supplement 1). We calculated the FF by counting spikes in sliding windows of 10 ms for each cell and each of the five bins (Figure 6). Confidence intervals for the FFs were obtained by resampling 1000 times. We also computed FFs using a Poisson process with absolute refractory period, with average touch response matched to the data (Berry II and Meister, 1998) (Figure 2—figure supplement 1).

Bottom Line: The amount of irreducible internal noise in spike trains, an important constraint on models of cortical networks, has been difficult to estimate, since behavior and brain state must be precisely controlled or tracked.The variance of touch responses was smaller than expected from Poisson processes, often reaching the theoretical minimum.Layer 4 spike trains thus reflect the millisecond-timescale structure of tactile input with little noise.

View Article: PubMed Central - PubMed

Affiliation: Janelia Research Campus, Howard Hughes Medical Institute, Ashburn, United States.

ABSTRACT
Cortical spike trains often appear noisy, with the timing and number of spikes varying across repetitions of stimuli. Spiking variability can arise from internal (behavioral state, unreliable neurons, or chaotic dynamics in neural circuits) and external (uncontrolled behavior or sensory stimuli) sources. The amount of irreducible internal noise in spike trains, an important constraint on models of cortical networks, has been difficult to estimate, since behavior and brain state must be precisely controlled or tracked. We recorded from excitatory barrel cortex neurons in layer 4 during active behavior, where mice control tactile input through learned whisker movements. Touch was the dominant sensorimotor feature, with >70% spikes occurring in millisecond timescale epochs after touch onset. The variance of touch responses was smaller than expected from Poisson processes, often reaching the theoretical minimum. Layer 4 spike trains thus reflect the millisecond-timescale structure of tactile input with little noise.

No MeSH data available.


Related in: MedlinePlus