Repository containing scripts, notebooks and preprocessed datasets for:

============================================================================

Patch-seq of endocrine cell exocytosis links physiologic dysfunction in diabetes to single-cell transcriptomic phenotypes

by Camunas-Soler J, Dai X, et al., https://doi.org/10.1101/555110

Description of Repository

• data folder: contains preprocessed datasets (gene count tables) for patch-seq dataset, FACS dataset, as well as electrophysiology and cell metadata. This folder also contains scRNAseq datasets from previous publications used in QC figures, as well as siRNA knockdown results.

  • Large files are compressed. Before runnning notebooks unzip tar files using tar -xvzf filename.tar.gz
  • Download data count matrix from Segerstolpe. et al dataset and unzip in data folder. Data can be found in Array Express Segerstolpe et al., file named E-MTAB-5061.processed.1.zip. After unzipping make sure Sandberg_pancreas_refseq_rpkms_counts_3514sc.txt is found in data folder.

• analysis folder: contains several folders with analysis results used to produce final figures. Data is saved in csv / excel files or pickles for integration with notebooks. Includes:

  • Cell typing file for each dataset.
  • tSNE coordinates for plots used in manuscript (patch-seq, patched vs non-patched, cryopreserved cells, alpha cells)
  • Machine Learning model for cell type classification based on Electrophysiology.
  • Correlations of gene expression to total exocytosis for beta-cells (nondiabetic, and T2D donors) and pathway analysis (compressed folder that needs to be unzipped).
  • Gene Set Enrichment Analysis (GSEA) for genes correlated to each functional group (i.e. Exocytosis, Sodium currents)
  • Correlations using a subset of beta-cells (train/test split) to determine Predictive Set of genes and perform predictions of electrophysiology.

• notebooks: Python and R notebooks to generate figures from manuscript and supplementary notebooks to generate results in analysis folders.

• functions: Helper functions to run notebooks.

• figures: Figures produced in notebooks.

• resources: databases and resources used in notebooks. Before runnning notebooks unzip compressed data files using command:tar -xvzf filename.tar.gz .