# Comparative microelectrode array dataset of functional development of human pluripotent stem cell-derived and rat neuronal networks 

## Authors
Fikret Emre Kapucu, Andrey Vinogradov, Tanja Hyvärinen, Laura Ylä-Outinen, Susanna Narkilahti

Neuro Group, Faculty of Medicine and Health Technology, Tampere University

## Summary

We provide a dataset of microelectrode array recordings (MEA) from human pluripotent stem cell (hPSC)-derived and rat embryonic cortical neurons during in vitro maturation and pharmacological treatment at their mature stage ([Kapucu et al., submitted](http://doi.org/to-be-inserted-after-publication)). Together with the recorded raw MEA data, we share the analysis code that produces the key scientific findings published previously with this dataset ([Hyvärinen et al., 2019](https://doi.org/10.1038/s41598-019-53647-8)), including PCA analysis to reveal the distinction between rat and human cell activity features. The data enables evaluation and comparison of hPSC-derived and rat cortical cultures in terms of functional development and response to pharmacological treatment. The 12500 Hz sampled data is recorded with Axion 12-well plates (for developmental tracking) and 48-well plates (for pharmacological experiments). Raw data constitute time series of extracellularly recorded voltage values in Volts. The duration of the recordings varies from 10 to 30 minutes depending on the experimental set-up. Each raw MEA data file contains organised meta-data. This dataset is valuable for experimental and computational neuroscientists and any signal analyst with interest for neuronal signalling. The accurate detailed description of data is delivered in the corresponding publication ([Kapucu et al., submitted](http://doi.org/to-be-inserted-after-publication)). Please, refer to the publication, as this README.md file deliver only a brief introduction of the data and basic procedures of reading it.

## Downloading the data

There are several ways to upload the data using GIN command line client, git annex and web browser. Short instructions for web browser and GIN CLI are provided below.

### Web browser
The clear way to get a particular file from the repository with the web interface is to proceed to its folder, click the file and then click the "Download" link below it. The files containing raw MEA signals are quite large, so the downloading time may be extensive in case of slow Internet connection. The robustness of the connection is important, as it happes so, that browsers sometimes can not continue the file downloading after abrupt connection loss.

### GIN CLI

Create an account on GIN and download the GIN client [(instructions)](https://gin.g-node.org/G-Node/Info/wiki/GIN+CLI+Usage+Tutorial). Log in using CLI:
```bash
gin login
```

Clone the repository:
```bash
gin get NeuroGroup_TUNI/Comparative_MEA_dataset
```

Big data files are not downloaded automatically. The following command should be used to get their contents:
```bash
gin get-content <filename>
```

Downloaded big files will be locked. You have to unlock the file with the following command:
```bash
gin unlock <filename>
```

To remove the contents of a big file: 
```bash
gin lock <filename>
gin remove-content <filename>
```

Please, refer to the detailed GIN CLI [Usage Tutorial](https://gin.g-node.org/G-Node/Info/wiki/GIN+CLI+Usage+Tutorial).

## Repository contents

### Directory Data
Data directory embeds the following sub-directories
* hPSC_MEA1, hPSC_MEA2, Rat_MEA1 <- developmental tracking
* hPSC_MEA3_Pharmacology, Rat_MEA2_Pharmacology <- pharmacological experiments
* PCA <- data sets for MEA plates, used in PCA analysis

PCA folder contains 7 sub-folders with pre-selected MEA sets used in PCA analysis (3 timepoints for each set): *hPSC_MEA1_PCA*, *hPSC_MEA2_PCA*, *hPSC_MEA4_PCA*, *hPSC_MEA5_PCA*, *Rat_MEA1_PCA*, *Rat_MEA3_PCA*, *Rat_MEA4_PCA*. PCA folder also includes *PCA_table.xlsx* file which contains extracted *meaRtools* output features of the sets for running MATLAB PCA analysis code (may be found in the Codes directory : /Codes/PCA_code/PCA.m).

Each folder contains raw MEA recordings in HDF5 format (used, for example as an input for MATLAB spike detection). Each .h5 file accomodate two groups: Data and DataInfo. Data group embeds subgroups for each well of MEA plate. Each of these subgroups contains a number of datasets for each electrode in the well. For example, to access the recording of electode 11 of well B3 of ... .h5 in MATLAB, the following command should be executed:

*h5read('...path/hPSC_20517_MEA1_DIV3.h5', '/Data/B3/11')* 

*DataInfo* group provides meta-data for the recording. It contains two datasets: *ExcludedWells* and *InactiveChannels* and a list of attributes (*SamplingFrequencyInHz*, *DurationInSec*, *RecordingUnits*, *DIV*, and *Plate type* which together deliver the detailed description of the raw MEA file. To load the dataset of excluded wells the following  command should be exectuted:

*h5read('...path/hPSC_20517_MEA1_DIV3.h5', '/DataInfo/ExcludedWells')* 

The attribute reading command is a bit different. For example, to read MEA plate type of the same file, the following command works: 

*h5readatt('...path/hPSC_20517_MEA1_DIV3.h5', '/DataInfo', 'Plate type')* 

Each folder also contains two sub-folders:
* *\<data label\>_spikes_noise_explogs* <- the folder with detected spike .csv files (MATLAB outputs) and explog and noisy electrodes .csv files (which all together comprise meaRtools input). 
* *\<data label\>_meaRtools_output* <- the folder with selected meaRtools outputs (those, used in *Hyvärinen et al., 2019*)

Pharmacology folders contain separate sub-folders for each stage of the pharmacological experiment (Baseline Pharmacology and TTX treatment). To check which well of MEA plate has undergone which specific treatment (e.g. drug or control) during Pharmacology and TTX stage it is recommended to take a glance into explog .csv file (to be found in the corresponding *\<data label\>_spikes_noise_explogs* sub-folder).



### Directory Codes
Codes directory contains the following folders:
* MATLAB spike detection
* meaRtools 
* Connectivity Analysis
* PCA code 

The implementational details can be found in the original publication.

## Related Publications
* Kapucu, F.E., Vinogradov A., Hyvärinen T., Ylä-Outinen L., Narkilahti S. Comparative microelectrode array dataset of functional development of human PSC-derived and rat neuronal networks. Submitted
* Hyvärinen, T., Hyysalo, A., Kapucu, F.E., Aarnos, L., Vinogradov, A., Eglen, J.E., Ylä-Outinen, L., Narkilahti, S. (2019). Functional characterization of human pluripotent stem cell-derived cortical networks differentiated on laminin-521 substrate: comparison to rat cortical cultures. Scientific Reports volume 9, 17125. https://doi.org/10.1038/s41598-019-53647-8

## Funding
* Academy of Finland, 332693 (Fikret Emre Kapucu), 286990/326436 (Laura Ylä-Outinen), 312414 and 311017 (Susanna Narkilahti)
* Orion Research Foundation, grant for postdoctoral research 2019 (Fikret Emre Kapucu), grant for doctoral dissertation work 2020 (Andrey Vinogradov)


## License
<a rel="license" href="https://creativecommons.org/licenses/by/4.0/"><img alt="Creative Commons License" style="border-width:0" src="https://licensebuttons.net/l/by/4.0/88x31.png" /></a><br /><span xmlns:dct="http://purl.org/dc/terms/" property="dct:title">Comparative microelectrode array dataset of functional development of human pluripotent stem cell-derived and rat neuronal networks</span> in the directory `Data` by <span xmlns:cc="http://creativecommons.org/ns#" property="cc:attributionName">Neuro Group, Faculty of Medicine and Health Technology, Tampere University (TUNI), Tampere, Finland</span> is licensed under a <a rel="license" href="http://creativecommons.org/licenses/by/4.0/">Creative Commons Attribution 4.0 International License</a>.

The directory `Codes\MATLAB spike detection` contains folder `Functions from F. Lieb's SpikeDetection-Toolbox` which embeds functions imported from [SpikeDetection-Toolbox](https://github.com/flieb/SpikeDetection-Toolbox) distributed under GNU General Public License v3.0. 

The directory `Codes` contains folder `meaRtools` with modified version of [meaRtools package](https://github.com/igm-team/meaRtools) distributed under GNU General Public License v3.0. Modifications imply user interface, input file reading and selection process and the addition of [LogISI algorithm](https://doi.org/10.1007/s10827-009-0175-1) for burst detection into the package. The LogISI algorithm originally received from its author Valentina Pasquale was modified for our data and the modifications are described in the publication [Kapucu et al., submitted](http://doi.org/to-be-inserted-after-publication). 

The directory `Codes` contains folder `Connectivity Analysis` with [Spectral entropy based neuronal network synchronization analysis: CorSE](https://doi.org/10.3389/fncom.2016.00112) code distributed under the 3-Clause BSD License.

All remaining codes in the directory `Codes` are published under the 3-Clause BSD License.

See the `LICENSE.txt` or `LICENSE` files in the corresponding directories for the full license.