Repository files navigation

Jellyfisher is an R package for visualizing tumor evolution and subclonalcompositions using Jellyfish plots, which display both spatial and temporaldimensions in a single unified figure.

The package is based on theJellyfish visualization tool,bringing its functionality to R users. Jellyfisher supports bothClonEvol results and plain data frames,making it compatible with various tools and workflows.

install.packages("jellyfisher")

devtools::install_github("HautaniemiLab/jellyfisher",build_vignettes=TRUE)

Jellyfisher is designed to work with data frames or ClonEvol results.

The input data should follow specific structures forsamples,phylogeny, andsubclonalcompositions, which are described in thejellyfisherfunction's documentation.

library(jellyfisher)# Plot the bundled example datajellyfisher(jellyfisher_example_tables)

Jellyfisher provides a straightforward way to visualizeClonEvol results using theextract_tables_from_clonevolfunction. See the reference page for details.

library(clonevol)library(jellyfisher)# Run ClonEvol. Check the ClonEvol documentation for details.y<- infer.clonal.models(...)y<- convert.consensus.tree.clone.to.branch(y)# Plot the resultsextract_tables_from_clonevol(y,model=1)|>  jellyfisher()

By default, all samples that have no explicit parent are children of theinferred root sample. You can customize the parent-child relationships bymodifying theparent column in thesamples data frame before plotting.

You can also modify the relationships with ease using theset_parentsfunction.

For example, if you have three samples, A, B, and C, they will have thefollowing relationships by default:

    Root   / | \  A  B  C

With the explicit parents, you can customize the relationships:

tables|>  set_parents(list(# The parent of C is B"C"="B"  )|>  jellyfisher()

    Root   /  \  A    B        \         C

Jellyfisher is a thin wrapper for theJellyfish visualization tool.Jellyfish is included as a git submodule in thetools/ directory.

To build the Jellyfish JavaScript dependency, run theupdate-and-build.shscript in thetools/directory. Most of the R code is autogenerated from the Jellyfish JavaScriptcode using thegenerate-R-code.mjsscript, which should be run after building the Jellyfish dependency.

If you use Jellyfisher in your research, please cite the following paper:

Kari Lavikka, Altti Ilari Maarala, Jaana Oikkonen, Sampsa Hautaniemi, Jellyfish:integrative visualization of spatio-temporal tumor evolution and clonaldynamics,Bioinformatics, 2025;, btaf091,https://doi.org/10.1093/bioinformatics/btaf091

Copyright (c) 2025 Kari Lavikka. MIT license, seeLICENSE.mdfor details.

Jellyfisher is developed inThe Systems Biology of Drug Resistance inCancergroup at theUniversity of Helsinki.

This project has received funding from the European Union's Horizon 2020research and innovation programme under grant agreements No. 965193(DECIDER) and No. 847912(RESCUER).