PICTograph is a computational tool developed to infer the clonal evolution of tumors from multi-region sequencing data. It models uncertainty in assigning mutations to subclones using a Bayesian hierarchical model and reduces the space of possible evolutionary trees by using constraints based on principles of sample-presence, lineage precedence, and sum condition. Highly probable evolutionary relationships that are recovered in multiple models can be highlighted using ensemble-based visualizations.
PICTograph uses the JAGS library for Bayesian data analysis, which is installed outside of R. JAGS can be downloaded and installed for your OS here.
To install PICTograph from GitHub, start R and enter:
devtools::install_github("KarchinLab/pictograph", build_vignettes = TRUE)
Demo code on a toy example can be found in the vignette:
library(pictograph)
vignette("pictograph", package = "pictograph")