Run tutorials on Google Colab

Wish to try Scarf without installation?

Google Colab allows running Python code directly on Google’s server through a notebook interface. With the following links you can try running any of the vignettes on Colab. This is a quick way to try Scarf without installing it.

Before you run notebooks on Colab

Paste the following code on the top of the notebook before running any other cell on Colab notebook

!pip install ipython-autotime
!pip install scarf
!pip install -U numpy scipy

Google Colab has older versions of Numpy and Scipy which are not compatible with Scarf. Once scipy and numpy have updated you will see a RESTART RUNTIME button. Click on it to activate latest versions. Now you are free to execute rest of the notebook.

Colab links

Basic pipelines

• Workflow for scRNA-Seq data

• Workflow for scATAC-Seq count matrices

Multi-omics/Multimodal analysis

• Analysis of Transcriptome + Surface Proteome

Data integration tutorials

• Projection of cells across datasets

• Merging datasets and partial training

Trajectory analysis tutorials

• Estimating pseudotime ordering and expression dynamics

Other Vignettes

• Understanding how data is organized in Scarf

• Getting data in and out of Scarf

• Cell subsampling using TopACeDo

• Demonstrating Scarf on MNIST dataset