shinySTRegister

The goal of shinySTRegister is to provide a simple (if barebones) interface for registering (i.e. aligning) two spatial transcriptomics samples. This is done by specifying three points on the first sample (to be aligned) and their corresponding projections on the second (reference) sample.

The corresponding affine transformation is then determined and applied to the first sample, which can be saved as RDS for future use.

Installation

You can install the development version of shinySTRegister from GitHub with:

# install.packages("devtools")
devtools::install_github("JonSulc/shinySTRegister")

Example

Given two samples (to_align and reference) loaded as SpatialExperiment objects, the interface can be started with:

library(shinySTRegister)
shiny_register(to_align, reference, nbins = 400)

The nbins parameter controls the resolution of the image, counts being binned into a grid of $nbins \times nbins$ pixels. This only affects the display resolution, the transformation being applied to the raw data coordinates.

Convenience functions

To help with the setup for this, there are a few convenience functions for the loading and display of data from certain technologies. load_as_spatial_experiment takes a counts filename (spot/cell $\times$ feature) and a spot/cell metadata filename or data.frame to create the SpatialExperiment object. The id_columns argument specifies the columns identifying the corresponding spots or cells to combine the data. If NULL, they are assumed to be in the