This is a BEAST2 package for Bayesian inference of molecular data for cancer evolution. This package implements error models and substitution models for inference of timed trees in a Bayesian MCMC framework.
This package requires Java 8, and at least
- BEAST 2 v2.6.6
- BEAST Labs v1.9.7
See dependencies in version.xml
The current release has the following features.
Error Models
- GT16 diploid nucleotide error model (16 states)
- Nucleotide error model (4 states - A, G, C, T)
- General error model (n states)
- Binary error model
Substitution Models
- GT16 diploid nucleotide model (16 states)
- Binary substitution model
- SiFit2 substitution model (2 states)
- SiFit3 substitution model (3 states)
You may install the Phylonco package using the Package Manager.
Start BEAUti, open the Package Manager by selecting File -> Manage packages from the Menu.
Click Package repositories to open a new popup window
Click Add URL and add "https://raw.githubusercontent.com/CompEvol/CBAN/master/packages-extra.xml" to the entry.
Click Done. The Phylonco package should now appear in the Package Manager
From the Package Manager, select Phylonco and click Install/Upgrade to install.
All models accept input genotypes in FASTA and Nexus formats. See here for the genotype codes.
For the GT16 model, VCF files can be converted to FASTA format using www.github.com/bioDS/vcf2fasta
Start the BEAST software
Make sure the "Use BEAGLE library" box is unchecked
Set the input file to one of the examples e.g., examples/test_GT16_error.xml
Launch beast in java only mode by adding the -java option, e.g.
Windows
java -jar c:\Users\BEASTUser\Desktop\BEAST\lib\launcher.jar -java beast.xml
Mac
/Applications/BEAST\ 2.6.6/bin/beast -java beast.xml
Linux
~/beast/bin/beast -java beast.xml
BEAUti UI is currently only supported for the Binary Substitution model. We are currently working on this :)
Tutorials and new models will be added once Beauti support is completed.
To build use ./gradlew clean build in the root directory, for details see dev notes
For manual package installation instructions, see here
-
Bouckaert at al. (2019). BEAST 2.5: An advanced software platform for Bayesian evolutionary analysis.
-
Chen et al. (2022). Accounting for errors in data improves timing in single-cell cancer evolution.
-
Kozlov et al. (2022). CellPhy: accurate and fast probabilistic inference of single-cell phylogenies from scDNA-seq data.
-
Zafar et al. (2017). SiFit: inferring tumor trees from single-cell sequencing data under finite-sites models.
-
Hou et al. (2012). Single-cell exome sequencing and monoclonal evolution of a JAK2-negative myeloproliferative neoplasm.