https://doi.org/10.5061/dryad.qv9s4mwn4

This dataset contains the raw and processed data derived from single-unit auditory nerve fibre recordings in young-adult and aging gerbils. Using high-impedance, glass electrodes, raw voltage traces and spike times were recorded from over one-thousand single fibres of the auditory nerve of Mongolian gerbils raised in a quiet environment. Software was written and tested in MATLAB version R2023b.

Description of the data and file structure

All raw, processed, and metadata of one animal is saved in one datafile ('G#.mat'). Such a .mat file contains one variable, 'exp', which is structured equally between all experiments. The content of each field in this struct is described in the document 'Data_Structure.pdf'. An overview of the metadata of each animal, e.g. age, hearing sensitivity, presented complex acoustic stimuli, number of single fibres, etc., can be found in 'metadata.csv'. Furthermore, there is one file with all animals combined into one struct called 'all_AN_data.mat'. In this file, the raw waveforms are removed to make the file size manageable. This file can be used to plot the outcomes of the analyses, e.g. best frequency vs. threshold of all fibres. For more details, please see the accompanying Data Descriptor manuscript.

Code to help search through the dataset

Three scripts are provided to help the user search through the full dataset as well as within a struct of one animal. check_dataset_metadata.m loops through the 'all_exp' struct and focusses on the metadata of the experiments. It recreates the metadata sheet, which can be used to select an animal of interest and investigate it further in check_dataset_animal.m. check_dataset_units.m loops through the 'all_exp' struct and focusses on the analysed outcomes of the single units. It can be used to plot any of these outcomes against each other, typically with the fibre’s best frequency on the horizontal axis. check_dataset_animal.m loops through the units of an ‘exp’ struct of one animal. It generates a scatterplot of the threshold against the best frequency of all the fibre’s recorded in that animal and a plot with all rate-level functions of that animal plotted in one graph. It also calls the function check_AN.m, which plots of a given recording the unfiltered first trial, the inter-spike interval histogram, the first 300 spike waveforms, and the median spike waveform +/- 95% confidence interval. The script also calls the function makePSTH.m, which is used to generate a peri-stimulus time histogram (PSTH) of all responses to tone bursts at or close to a given stimulus level above the fibre’s threshold.

Code to re-analyse the spike times for auditory nerve fibre characterization

The script call_extract_func.m calls all functions that were used to generate the analysed outcomes of one unit of one animal. These include the following functions:

1.     BFextract_func.m, which generates the frequency-response curve to derive best frequency.

2.     CFextract_func.m, which generates the receptive field and tuning curve to derive characteristic frequency, threshold, and Q10dB, a measure of frequency selectivity.

3.     PHextract_func.m, which calculates vector strength and plots it as a function of stimulus level.

4.     CLICKextract_func.m, which calculates click latency in three different ways and visually demonstrates the outcomes of each method.

5.     RLFextract_func.m, which generates the rate-level function to derive threshold and rates at each stimulus level.

6.     SRextract_func.m, which calculates the spontaneous rate from a long recording in silence.

Sharing/Access information

We plan to make both the data and the software also available in the Auditory Modeling Toolbox v1.6 under data_heeringa2024 and exp_heeringa2024.