Stimulus background influences phase invariant coding by correlated neural activity
- MGill University, Canada
Figures
Figure 1
A chirp with given duration and frequency excursion gives rise to heterogeneous waveforms depending on time of occurrence within the beat cycle as well as beat frequency.
(A) Two weakly electric fish with their electric organ discharges (EODs) in red and blue. (B) The EOD waveforms of both fish (top red and blue traces) show alternating regions of constructive and destructive interference when the instantaneous EOD frequencies do not vary in time (middle red and blue traces). Interference between the EODs leads to a sinusoidal amplitude modulation (i.e. a beat, bottom black trace) of the summed signal (bottom brown trace). (C) Schematic showing communication between the emitter (red) and receiver (blue) fish. (D) During a communication call, the emitter fish’s EOD frequency (top red trace) transiently increased by a maximum of Δf for a duration Δt (top green trace) while the receiver fish’s EOD frequency (top blue trace) remains constant. The communication call results in a phase reset of the beat (bottom black trace). (E) Ongoing unperturbed beat (gray dashed traces) and stimulus waveforms (black traces) resulting when a chirp (green) with the same frequency excursion Δf and duration Δt occurs at different phases during beat cycles with frequencies 2 Hz (left), 16 Hz (middle), and 64 Hz (right). We note that the waveforms of the remaining four phases are mirror images of the waveforms shown. (F) Detectability of chirps occurring at different phases of the beat cycle as a function of beat frequency. Chirps occurring at higher beat frequencies are harder to detect as they are more similar to the beat. (G) Distance between chirp waveforms occurring at different phases of the beat cycle across beat frequencies. Inter-chirp distances remain constant across the beat frequencies used. ‘*' indicates statistical significance to all values obtained for lower frequencies at the p=0.05 level using a one-way ANOVA with Bonferroni correction.
-
Figure 1—source data 1
Source data for Figure 1.
- https://doi.org/10.7554/eLife.24482.003
Figure 2 with 2 supplements
Phase invariant coding of chirps by correlated but not single neuron activity is best at low and deteriorates for higher beat frequencies.
(A) Schematic showing the experimental setup. (B) Example stimulus waveforms (top) for chirps occurring at different beat frequencies (black: 2 Hz; dark gray: 16 Hz; light gray: 64 Hz), normalized firing rates (middle, blue), and spike count correlations (purple) from example afferent pairs. (C). Population-averaged firing rate responses (top, blue) and spike count correlation as a function of time (bottom, purple) to the different stimulus waveforms occurring at a beat frequencies of 2 Hz (left), 16 Hz (middle), and 64 Hz (right). Zero indicates the time at chirp onset. (D) Invariance score computed from single afferent activity (blue) and from correlated activity (purple) for all phases as a function of background beat frequency. (E) Detectability of chirp waveforms over the beat for population-averaged firing rate responses (blue) and spike count correlation (purple) as a function of background beat frequency. ‘*” indicates statistical significance from all values obtained for frequencies <64 Hz at the p=0.05 level using a one-way ANOVA with Bonferroni correction. We note that, in this and other figures, all neural responses are shifted to the left by 9 ms in order to account for known axonal transmission delays.
-
Figure 2—source data 1
Source data for Figure 2.
- https://doi.org/10.7554/eLife.24482.005
Figure 2—figure supplement 1
Responses to small chirps occurring on top of beats with varying frequencies.
(A) Raster plots of two example afferent responses (blue) and their correlated activity (purple) to chirps occurring at different phases for background beat frequencies of 2 Hz (left), 16 Hz (middle) and 64 Hz (right). The red boxes indicate the response duration for single unit firing rates and correlated activity, respectively, after chirp onset for each chirp-beat combination shown. Note that the remaining four chirp phases are mirror images of the examples shown. (B) Duration of single neuron spiking activity (blue) and of correlated activity (purple) in response to a chirp as a function of background beat frequency. Also shown are the best power paw fits with exponents and R2 values (C) Invariance scores for single unit firing rate using a window size computed from the response duration of primary afferents to chirps (filled) and from the response duration of correlated activity to chirps (hollow). (D) Detectability of chirp waveforms over the beat frequencies for population-averaged firing rate responses using a window whose size was taken from the response duration of primary afferents (filled) and from the response duration of correlated activity (hollow).
-
Figure 2—figure supplement 1—source data 1
Source data for Figure 2—figure supplement 1.
- https://doi.org/10.7554/eLife.24482.007
Figure 2—figure supplement 2
Phase locking in primary afferents to different background beat frequencies.
(A) Phase locking index as a function of background beat frequency. (B) Phase histograms of example afferents to background beat frequencies of 2 Hz (top), 16 Hz (middle) and 64 Hz (bottom). Note that complete cessation of firing (i.e. rectification) is only observed in response to the 64 Hz beat frequency.
-
Figure 2—figure supplement 2—source data 1
Source data for Figure 2—figure supplement 2.
- https://doi.org/10.7554/eLife.24482.009
Figure 3 with 1 supplement
Detectability and invariance of perception are best predicted by correlated afferent activity for different beat frequencies.
(A) Experimental setup. Each fish (N = 33) was placed in an enclosure within a tank (chirp chamber). Stimuli were applied via two electrodes (S1 and S2) perpendicular to the fish’s rostro-caudal axis. The fish’s EOD frequency was recorded by a pair of electrodes positioned at the head and tail of the animal (E1 and E2). Behavioral responses consisted of communication stimuli characterized by transient increases in EOD frequency in response to the presented stimulus. (B) Population-averaged time-dependent behavioral response rates in response to chirps occurring at different phases of the beat cycle for beat frequencies of 2 Hz (left), 16 Hz (middle), and 64 Hz (right). (C) Population-averaged behavioral invariance scores computed from behavioral responses obtained for different beat frequencies (brown) in comparison to the neuronal invariance scores using correlated activity (purple) and single units (blue). Note that the behavioral invariance (brown) across beat frequencies follows the one obtained for correlated activity (purple) but not single units (blue). (D) Chirp rate as a measure of echo response to the chirp waveforms played (brown) compared to chirp detectability computed for correlated activity (purple) and single unit firing rate (blue) as a function of background beat frequency. Note that the behavior (brown) matches the correlated activity (purple). ‘*' indicates statistical significance to all values obtained for frequencies <64 Hz at the p=0.05 level using a one-way ANOVA with Bonferroni correction.
-
Figure 3—source data 1
Source data for Figure 3.
- https://doi.org/10.7554/eLife.24482.011
Figure 3—figure supplement 1
Characteristics of real chirps do not vary across baseline beat frequencies.
(A) Probability distributions of the frequency increase of small chirps recorded at different baseline beat frequencies. All samples originate from the same distribution (Kruskal-Wallis test, df = 7; Chi2 = 0.38; p=0.9998) (B) Probability distributions of the duration of small chirps recorded at different baseline beat frequencies. Note that all samples originate from the same distribution (Kruskal-Wallis test, df = 7; Chi2 = 9.74; p=0.2038).
-
Figure 3—figure supplement 1—source data 1
Source data for Figure 3—figure supplement 1.
- https://doi.org/10.7554/eLife.24482.013
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Stimulus background influences phase invariant coding by correlated neural activity
eLife 6:e24482.
https://doi.org/10.7554/eLife.24482
