*library_books*

## Similar protocols

## Protocol publication

[…] For each stimulus presentation, we recorded neuronal activity, pulses time aligned with each frame of the stimulus, and vertical synchronization (Vsync) pulses from the video card. Neural activity was amplified with a differential AC amplifier (A‐M Systems, model No. 1700, Sequim, WA) with a gain of 10,000 and sampled at 25 kHz. To store the data, a RP2.1 enhanced real‐time processor (Tucker‐Davis Technologies, Alachua, FL) with Butterworth filter settings of 100 Hz (high pass) and 5 Hz (low pass) was used. The axon of the DCMD is the largest in the ventral nerve and produces characteristically large amplitude spikes (Fig. A), which were easily identified and isolated using threshold sorting in Offline Sorter (Plexon Inc., Dallas, TX). Spike times (Fig. B) were exported to **Neuroexplorer** (NEX Technologies, Littleton, MA) and used to create peristimulus time histograms (1‐msec bin) smoothed with a 50 msec Gaussian filter (Fig. C).Trajectories were grouped for analysis based on whether they included a time of collision (TOC), time of transition (TOT), or both (TOC/TOT). In trials where a subset of individual animals did not respond (i.e., no firing rate modulation) to a transition (i.e., H_T and 0_90_T), only responders were used in statistical analysis (n = 18 and 14, respectively). To characterize DCMD firing properties, we used similar methods to those described in McMillan and Gray () and Silva et al. (). For all trajectories, we measured the total spike number during the entire stimulus presentation as well as the time leading up to transition. For all trajectories that contained a looming component (L_0, H_T, L_45, and 0_90_T), we quantified the positive peak firing rate (f
TOCp) and peak time (t
TOCp) associated with the projected TOC. We also quantified the negative peak (valley) firing rate (f
TOTv) associated with TOT for trajectories that transitioned toward looming (H_T, 0_90_T) and the positive peak firing rate (f
TOTp) associated with TOT for trajectories that transitioned away from looming (0_45, H_A, L_45_A, and 0_90_A). For transitions, we also measured the time of the respective valleys (t
TOTv) or peaks (t
TOTp
) relative to TOT. For trajectories H_A, H_T, and 0_90_T, TOT‐associated responses were masked by compression of the response duration in a flow field, and thus we were unable to discern a TOT‐associated valley or peak for this background. For trajectories with collision phases, we measured peak width of the peristimulus time histogram at ½ the maximal firing rate. The rise phase duration(s) of each DCMD response was calculated from the time at which the DCMD firing rate surpassed a 99% confidence interval (calculated from the entire stimulus presentation) to the time of peak DCMD firing rate (Fig. C). The TOC‐associated fall phase duration was designated as the time between when the stimulus stopped expanding and when the firing rate decreased to 15% of the peak DCMD firing rate (Gabbiani et al. ; Guest and Gray ). For trajectories that deviated from looming, TOT‐associated fall phases were designated as the duration from the maximum firing rate of the TOT‐associate peak to when the firing rate decreased to 15% of this value. For trajectories that deviated to looming, the maximum TOT‐associated firing rate occurred at TOT and thus the falling phase was equal to the delay (δ) from TOT to the time of the valley and, therefore, these data are equal to t
TOTv for trajectories H_T and 0_90_T. Previous studies describe how the rise and fall phases relate to stimulus‐evoked presynaptic network activity that drives LGMD/DCMD responses to motion (Gabbiani et al. , ; Silva et al. ). For example, the fall phase relates to feed‐forward inhibition that terminates responses to looming (Gabbiani et al. ).We also quantified object expansion parameters known to correlate with DCMD firing rate modulation (f′) and the delay from TOT to the time of the resulting peak or valley (δ): the instantaneous acceleration of: (1) the looming stimulus angular subtense (θ); and (2) the angular motion of the leading edge (ψ). Our data were pooled with those from McMillan and Gray (), Dick and Gray () and Silva et al. () and fit to unconstrained two‐dimensional Gaussian equations of the form:(1)f′=3.1e−0.5θ″+3.833.32+ψ″+166.8123.62
(2)δ=0.1e−0.5θ″+4.638.52+ψ″+352.2751.42
where θ″ is the instantaneous acceleration of θ, ψ″ is the instantaneous acceleration of ψ, 3.1 and 0.1 represent the height of the mesh plot for f′ and δ, respectively, the numerical values within each successive numerator define the center of the peak of the mesh plot and the successive denominators relate to the width of the curve in the x and y planes, respectively. [...] Firing parameters were tested for normality and equal variance in response to different trajectories and backgrounds with **SigmaPlot** 12.5 (Systat Software, Richmond, CA). DCMD firing parameters between initial and final frontal looms were compared using a t‐test for parametric data or a Mann–Whitney Rank Sum test for nonparametric data. The full data set did not satisfy tests for normality or equal variance. Therefore, we tested for significant differences across trajectories or backgrounds using a one‐way ANOVA on Ranks (reported by the H statistic). Flow fields resulted in no responses to a trajectory change for H_A and H_T and, therefore, data from simple and scattered backgrounds were compared using a Mann–Whitney Rank Sum test (reported by the U statistic). We used either a Tukey or Dunn's pairwise post hoc comparison for data with equal or unequal sample sizes, respectively. Table summarizes the results of all statistical tests. All data were plotted as box plots showing the median value, 25th and 75th percentile as box boundaries, 10th and 90th percentiles as error bars, and outliers as small filled circles. […]

## Pipeline specifications

Software tools | NeuroExplorer, SigmaPlot |
---|---|

Applications | Miscellaneous, Neurophysiology analysis |

Diseases | Macular Edema |