Why (interactive) visualization?

What is iSEE?

• An open-source R/Bioconductor package that allows you to easily create customized interactive interfaces for exploring your data and visualize any pre-calculated results.

What is iSEE?

• An open-source R/Bioconductor package that allows you to easily create customized interactive interfaces for exploring your data and visualize any pre-calculated results.

• Essentially data type agnostic - any rectangular data can be explored (but it was designed with high-throughput biological data in mind).

• Easy to run (for a suitably formatted data set, more on that shortly):

  library(iSEE)
  iSEE(sce)

• Highly configurable, flexible interface.

• Strong focus on reproducibility, all code to generate displayed figures can be downloaded.

What about other tools?

https://github.com/federicomarini/awesome-expression-browser

With what data can iSEE be used?

• Any form of (numeric) rectangular-shaped data (measurements for some form of features across a range of samples).
• Both features and samples can come with annotations/additional information.
• Data must be stored in a SummarizedExperiment container (or one of its derivatives, including SingleCellExperiment or DESeqDataSet).

With what data can iSEE be used?

• Any form of (numeric) rectangular-shaped data (measurements for some form of features across a range of samples).
• Both features and samples can come with annotations/additional information.
• Data must be stored in a SummarizedExperiment container (or one of its derivatives, including SingleCellExperiment or DESeqDataSet).

How to get to a SummarizedExperiment/SingleCellExperiment object?

• Use the constructor functions: SummarizedExperiment(), SingleCellExperiment(), DESeqDataSet(), DESeqDataSetFromMatrix(), ...

• Many Bioconductor packages directly import data into a SingleCellExperiment object (e.g., DropletUtils::read10xCounts()) or a SummarizedExperiment object (e.g., tximeta::tximeta()).

• From an AnnData object: use e.g. the zellkonverter Bioconductor package.

• From a Seurat object: use e.g. Seurat::as.SingleCellExperiment().

Default panels

Some additional iSEEu panels (1)

Some additional iSEEu panels (2)

Some additional iSEEu panels (3)

Some additional iSEEu panels (4)

Additional resources

• The bookdown book on Extending iSEE: https://isee.github.io/iSEE-book/
• Bioconductor landing page: https://bioconductor.org/packages/iSEE/
• Publication (F1000Research, 2018): https://f1000research.com/articles/7-741/v1
• Deployed examples: https://marionilab.cruk.cam.ac.uk/, code at https://github.com/iSEE/iSEE2018
  • Further deployments in the https://github.com/iSEE/iSEE_instances repo
  • iSEE in production: http://www.teichlab.org/singlecell-treg, https://libd.shinyapps.io/tran2020_Amyg/, http://shiny.imbei.uni-mainz.de:3838/iSEE_covidIT/
• Development version (bug reports etc): https://github.com/iSEE/iSEE
• The #iSEE channel in the Bioconductor slack workspace
• Additional panels and modes in iSEEu: https://bioconductor.org/packages/iSEEu/

Kevin Rue-Albrecht, Federico Marini, Charlotte Soneson, Aaron Lun