Tools for simulating breast tumor heterogeneity and calculating heterogeneity metrics using both simulated and clinical data
Description
heterogeneityMetrics.R is an R script that can be used to calculate relevant tumor heterogeneity metrics using mutation calls from multiregion sequencing (MRS) of human tumor specimens (using the output of the VAP pipeline) and using mutation calls from simulated data (using the output of breastSimulationCode.py). All of the metrics except for tHFR take as input mutation calls from multiple regional samples from within the same timepoint. tHFR takes as input one set of "pre-treatment" mutations as well sets of mutations from multiple "post-treatment" regions. These heterogeneity metrics can be used to measure genetic divergence between regions and test for clonal evolution across time (tHFR).
breastSimulationCode.py is a Python script to simulate 3D peripherally dominated tumor growth and output multi-region sequencing data (mutation calls) via an agent-based model. Deme subdivision is assumed in order to model cell mixing and spatial constraint. 20 samples across all octants of the deme are recorded to profile simulated spatial tumor heterogeneity. This version assumes at most two drivers occur on the same lineage; the fitness of Tier1 and Tier2 driver lineages is 1+s and (1+s)^2, respectively.
Requirements
Python packages: numpy,sys,math,random,collections,sets
Usage
Simulation of a typical tumor (~10^9 cells) is computationally costly. We suggest to run this script on a high performance cluster. The memory cost is also large (about ~40G when the final_tumor_size = 10^9 and mut_rate = 0.6).
To run the python script
$ python breastSimulationCode.py s_coef repl
e.g., python breastSimulationCode.py 0.1 2
Since breast tumors follow patterns of heterogeneity matching growth under strong selection, we suggest simulated with selection coefficients (s_coef up to 0.4-0.5). Deme size can also be modified within the script, as can the number of public mutations and mean sequencing depth for the outputted mutations. (edited)
Contact
Katherine Lee Pogrebniak: kpogrebn@stanford.edu