Application of a Variational Autoencoder for Clustering and Analyzing in situ Articular Cartilage Cellular Response to Mechanical Stimuli

This is the dataset for our paper titled " Application of a Variational Autoencoder for Clustering and Analyzing in situ Articular Cartilage Cellular Response to Mechanical Stimuli".

Corresponding author information:

Email: [e-mail] (Han Kheng Teoh); [e-mail] (Itai Cohen)

This dataset is shared under a Creative Commons Attribution 4.0 International license (CC BY 4.0); the data will be openly available to share and adapt, but appropriate credit to the original data creators is required upon reuse.

When using this dataset, please cite:

The dataset:

Jingyang Zheng, Han Kheng Teoh, Michelle L. Delco, Lawrence J. Bonassar , and Itai Cohen. (2024) Data from: Application of a Variational Autoencoder for Clustering and Analyzing in situ Articular Cartilage Cellular Response to Mechanical Stimuli [dataset]. Zenodo. https://doi.org/10.5281/zenodo.10565588

AND the paper: 

Jingyang Zheng, Han Kheng Teoh, Michelle L. Delco, Lawrence J. Bonassar , and Itai Cohen. (2024) Application of a Variational Autoencoder for Clustering and Analyzing in situ Articular Cartilage Cellular Response to Mechanical Stimuli. PLOS One  https://doi.org/10.1371/journal.pone.0297947

The work was supported by the NIH National Institute of Arthritis and Musculoskeletal and Skin Diseases, Contract: K08AR068470, R03AR075929, and The Harry M. Zweig Fund for Equine Research. This work was also supported by the NIH National Institute of Neurological Disorders and Stroke. Contract: R01NS116595. Additionally, this work was supported by the National Science Foundation grants DMR-1807602, CMMI 1927197, and BMMB-1536463. Lastly, this work made use of the Cornell Center for Materials Research Shared Facilities, which are supported through the NSF MRSEC program (DMR-1719875).

DATA & FILE OVERVIEW

-------------------------------------------------

The dataset contains two folders : Data and Code.

In the Data folder, the experimental data is organized into three subfolders, specifying the date when the experiment was performed.

Each subfolder contains the following files:

all_locs.mat - contains the cells (x,y) position. The data is organized as a Nx2 array, where N is the number of cells in the sample. 

blue_all.mat - contains the post-impact NMP (cell death) intensity for each cell. The data is organized as a TxN array, where T is the number of time points the NMP (cell death) intensity was measured.

green_all.mat - contains the post impact Ca^{2+} intensity for each cell. The data is organized as a TxN array, where T is the number of time points the Ca^{2+} intensity was measured.

red_all.mat - contains the post impact TMRM (mitochrondrial polarity) intensity for each cell. The data is organized as a TxN array, where T is the number of time points the TMRM (mitochondrial polarity) intensity was measured.

impact_intensity.mat - contains the Ca^{2+} intensity during impact for each cell. The data is organized as a TxN array, where T is the number of time points the Ca^{2+} intensity was measured.

impact_locs.mat - contains the cells' (x,y) position within the impact site. The data is organized as a Nx2 array, where N is the number of cells in the sample.  

D_skl_dd_mm_yy.p - contains the symmetrized KL divergence between cells' latent representation. The data is organized as a N by N array, where N is the number of cells.

In addition, the Data folder also contains :

model_weights.p file - contains the weights and biases for the trained VAE network used in this study.

The Code folder contains:

decoders.py -  contains a class function for the VAE decoder.

encoders.py -  contains a class function for the VAE encoder.

loaders.py -  contains a function that partitions the cell data into a training set and a test set.

wrapper.py - contains a class function that trains a VAE. 

Cartilage VAE - Part I.ipynb - contains the code necessary to generate Figures 1 to 5 in the manuscript.

Cartilage VAE - Part II.ipynb - contains the code necessary to generate Figures 5 to 9 in the manuscript.