This repository contains the code to reproduce plots presented in our BioVis talk+poster at ISMB '22. The talk itself is available on YouTube.
For a much more elaborate R implementation that includes our FAUST clustering method, please take a look at https://github.com/RGLab/FAUST.
For details about our clustering and visualization methods, please take a look at the related publication:
git clone [email protected]:flekschas-ozette/ismb-biovis-2022.git
cd ismb-biovis-2022
conda env create -f environment.yml
conda activate ozette-ismb-biovis-2022
Download the example data from https://figshare.com/articles/dataset/ISMB_BioVis_2022_Data/20301639 and place the files under data/mair-2022
. The data is from Mair et al., 2022, Extricating human tumour immune alterations from tissue inflammation, Nature.
-
Start JupyterLab:
jupyter-lab
-
Open one of the following notebooks:
-
Explanation of our transformation embedding approach: annotation-embedding.ipynb
-
Comparison of our transformation approach using different non-linear embedding methods: compare-annotation-embedding.ipynb
-
Joint embedding of two samples showing how our transformation approach helps to reduce batch effects: joint-annotation-embedding.ipynb