STRIDE

Spatial TRanscrIptomics DEconvolution by topic modeling (STRIDE), is a computational method to decompose cell types from spatial mixtures by leveraging topic profiles trained from single-cell transcriptomics. Besides the cell-type composition deconvolution, STRIDE also provides several downstream analysis functions, including (1) signature (i.e., topic) detection and visualization, (2) spatial clustering and domain identification based on neighborhood cell populations and (3) reconstruction of three-dimensional architecture from sequential ST slides of the same tissue.

Documentation

For full installation and usage of STRIDE, please refer to the documentation.

Change Log

v1.0.1

• Build STRIDE.

v1.0.2

• Add mapping function to identify similarest cells for spatial spots.
• Fix bugs of integration.
• Fix bugs of deconvolution with trained topic model.

v1.0.3

• Update mapping function to make cell mapping consistent with the deconvolved fractions.

v1.0.4

• Fix a bug that produced an empty *_spot_celltype_frac.txt when a topic column summed to zero (NaN propagated through the deconvolution). Zero-sum topic columns are now filled with 0 after normalization.

Install STRIDE

git clone https://github.com/DongqingSun96/STRIDE.git
cd STRIDE
pip install -r requirements.txt
python setup.py install

Usage

STRIDE --help
usage: STRIDE [-h] [-v] {deconvolve,plot,cluster,integrate,map} ...

STRIDE (Spatial TRanscrIptomics DEconvolution by topic modelling) is a cell-
type deconvolution tool for spatial transcriptomics by using single-cell
transcriptomics data.

positional arguments:
  {deconvolve,plot,cluster,integrate}
    deconvolve          Decompose celltype proportion for spatial
                        transcriptomics.
    plot                Visualize the deconvolution result.
    cluster             Neighbourhood analysis based on cell-type composition
                        and local cell population
    integrate           Integrate multiple samples from the same tissue.
    map                 Identify similarest cells for spatial spots.

optional arguments:
  -h, --help            show this help message and exit
  -v, --version         Print version info.