Matlab r2019a

Statistical analysis of the bat level of focus and call centroid distance, Fig. 3 h and i were conducted in RStudio (Version 1.2.5042, RStudio, Inc.) using the lme4 package by Bates et al.^{59 (link)}. Testing for normal distribution was performed with Shapiro-Wilk normality test. Data were analyzed by applying a non-parametric two-sided Wilcoxon signed-rank tests and we fitted a linear model. Data preparation and calculation of basic statistical parameters (mean; standard deviation; median) was performed in Microsoft Excel. Graphs were generated using MATLAB_R2019a, from Mathworks or in Rstudio using the ggplot package from Wickham⁶⁰ . The data that support these findings are available online⁶¹ .

Fluorescence microscopy setups were controlled by custom built software written in LabView 2016 for Windows (National Instruments, Austin, TX). Single-vesicle fusion data were analyzed using custom built software written in LabView 2016 for Windows (National Instruments) as described^{39 (link)}. FLIC intensities were extracted using custom software written in LabView 2016 for Windows (National Instruments, Austin, TX) and fitted using FLIC v.0.5, which was generously provided by Armin Lambacher and Peter Fromherz (Max Planck Institute for Biochemistry, Martinsried/München, D 82152 Germany)^{35 (link)}. The simulation of the fluorescent release lineshapes (Fig. 5c) was performed in Matlab_R2019a (Mathworks, Natick, MA). Fits of kinetic fusion data and assembly of data graphs were performed using Igor Pro 8 (Wavemetrics, Lake Oswego, OR). EPR spectra and depth parameters were plotted using OriginPro, versions 7.5 or 2021 (OriginLab, Northhampton, MA). Molecular Structures were rendered using PyMol (Schrodinger, NY, NY).

Considering the PHG is a substantial brain region with multifaceted functions (Aminoff et al., 2007 (link); Stenger et al., 2022 (link)), we segregated it into six subdivisions (Figure 1A) to explore the functions of each subdivision, respectively. These divisions, our regions of interest (ROIs), were based on the empirically validated Brainnetome Atlas (BNA) (Fan et al., 2016 (link)).²We used SPM8 software³ within MATLAB R2019a (MathWorks Inc., Natick, MA, USA) to compute each ROI’s mean fALFF values. For each voxel’s time series, we obtained the sum of the amplitudes within the low-frequency range (0.01–0.1 Hz). The fALFF values for each voxel and each participant were then calculated by dividing this fractional low-frequency amplitude sum by the amplitude sum across the total frequency range (0–0.25 Hz) (Zou et al., 2008 (link); Zuo et al., 2010 (link)). As a normalized ALFF index, fALFF minimizes the influence of artifactual signals near vessels or significant pulsatile motion (Zou et al., 2008 (link); Zuo et al., 2010 (link)). The fALFF value maps were registered to the MNI152 space using the previously computed transformation matrix. Furthermore, we estimated each participant’s head motion from the rs-MRI data by incorporating the head motion correction outputs to further minimize the head motion’s influence.