Cell-cell communications are vital for biological signaling and play important roles in complex diseases. Recent advances in single-cell spatial transcriptomics (SCST) technologies allow examining the spatial cell communication landscapes and hold the promise for disentangling the complex ligand-receptor interactions across cells. However, due to frequent dropout events and noisy signals in SCST data, it is challenging and lack of effective and tailored methods to accurately infer cellular communications. Herein, to decipher the cell-to-cell communications from SCST profiles, we propose a novel adaptive graph model with attention mechanisms named spaCI. spaCI incorporates both spatial locations and gene expression profiles of cells to identify the active ligandreceptor signaling axis across neighboring cells.

• spaCI incoporates both spatial locations and gene expressions of cells for revealing the active ligand-receptor signaling axis across neighboring cells.
• spaCI is able to identify the upstream transcriptional factors mediating the identified ligand-receptor interactions, which allows gaining further insights into the underlying cellular communications, the discoveries of disease mechanisms, and effective biomarkers.
• spaCI is developed tailored for spatial transcriptomics and provided available as a ready-to-use opensource software, which demonstrates high accuracy and robust performance over existing methods.

• For the spaCI tutorial and step-by-step manual, please refer to the jupyter notebook here
• We provide a toy demo with one-command bash script for parameter tuning in spaCI, please refer to here
• Toy data can be downloaded at here

Tang Z, Zhang T, Yang B, Su J, Song Q, spaCI: deciphering spatial cellular communications through adaptive graph model, Briefings in Bioinformatics, 2022;, bbac563, https://doi.org/10.1093/bib/bbac563

• How can I install spaCI?
  You can download spaCI from our github link:

  git clone https://github.com/QSong-github/spaCI.git
  

  spaCI is built based on pytorch, tested in Ubuntu 18.04, CUDA environment(cuda 11.2) the requirement packages includes:

  torchvision==0.11.1
  torch==1.6.0
  tqdm==4.47.0
  typing==3.7.4.3
  numpy==1.13.3
  pandas==1.5.1
  PyYAML==6.0
  

  or you can also use the following scripts:

  pip install -r requirements.txt
  

• I want to try the toy demo, can I run spaCI in one command line?
  You can use the following commands:

  python configuration.py --trainroot dataset/triplet.csv --testroot dataset/test_pairs.csv --predroot dataset/test_lr_pairs.csv --matrixroot dataset/exp_data_LR.csv --adjroot dataset/spatial_graph.csv --ymlname conf.yml --threshold 0.9' 
  python main_yaml.py
  

  or please refer to our training tutorial here

• How can I apply spaCI in my own dataset? And how to generate the desired format for spaCI?
  We prepare a tutorial for data prepraration here

• Do I need a GPU for running spaCI?
  spaCI is able to run on a standard laptop without GPU. For computational efficiency, we provide options in the configuration.yml file, with the setting use_cuda='cpu' or use_cuda='gpu'.

• Can I generate my own configuraton file using command line?
  To enable users generating their specific configure.yml with different hypterparameters, we provide a script to generate the configuration file. Please refer to here for details.