An R project developed to simulate somatic evolution in cancer, with a focus on the role of L1 elements

Dependencies:

Required:

• data.table
  • https://cran.r-project.org/web/packages/data.table/data.table.pdf
  • install.packages('data.table')
• parallel
  • https://stat.ethz.ch/R-manual/R-devel/library/parallel/doc/parallel.pdf
  • install.packages('parallel')

For run_v2_7.r script:

• optparse
  • https://cran.r-project.org/web/packages/optparse/optparse.pdf
  • install.packages('optparse')

For run_v2_7.ipynb notebook:

• Jupyter notebook
  • http://jupyter.org/install
• IRKernel
  • https://irkernel.github.io/installation/z

Source Code

./src/run_v2_7.r

• An R script for running sompop_v0_2_7 simulation.

Usage: ./run_v2_7.r [options]


Options:
	--drivers=DRIVERS
		List of driver genes: text file with single column of driver gene ENSEMBL IDs
		[default = ../data/gene_lists/drivers_lusc_filt.txt (Lung Squamous Cell Carcinoma)]

	-g GENDER, --gender=GENDER
		Gender of host: 'male' or 'female' [default = female]

	-n INITIAL_SIZE, --initial_size=INITIAL_SIZE
		Number of cells in initial population [default = 1000]

	-m INSERTION_RATE, --insertion_rate=INSERTION_RATE
		Average number of L1 insertions per cell cycle [default = 1]

	-D DRIVER_STRENGTH_HOM, --driver_strength_hom=DRIVER_STRENGTH_HOM
		Selection coefficient of homozygous driver mutations [default = 0.1]

	-P PASSENGER_STRENGTH_HOM, --passenger_strength_hom=PASSENGER_STRENGTH_HOM
		Selection coefficient of homozygous passenger mutations [default = 0.005]

	-d DRIVER_STRENGTH_HET, --driver_strength_het=DRIVER_STRENGTH_HET
		Selection coefficient of heterozygous driver mutations [default = DRIVER_STRENGTH_HOM * 0.1]

	-p PASSENGER_STRENGTH_HET, --passenger_strength_het=PASSENGER_STRENGTH_HET
		Selection coefficient of heterozygous passenger mutations [default = PASSENGER_STRENGTH_HOM * 0.1]

	-t NUMBER_TIMESTEPS, --number_timesteps=NUMBER_TIMESTEPS
		Number of time steps to simulate
		[default = 1000]
		(Note: by default 1 time step = 1 cell generation)

	--tau=TAU
		Time resolution: number of generations simulated per timestep (can be < 1) [default = 1]

	-l LOG_PATH, --log_path=LOG_PATH
		Path to log file [default = ./log.txt]

	--parallel=PARALLEL
		Whether to use parallel processing [default = FALSE]

	--initial_mut=INITIAL_MUT
		'none' - no mutations in initial population
		'single' - 1 gene is homozygously disrupted in a single cell of population (flag --initial_gene specifies this gene)
		'inherited' - 1 gene is heterozygously disrupted in all cells of initial population (flag --initial_gene specifies this gene)
		[default = none]

	--initial_gene=INITIAL_GENE
		Gene disrupted at initialization if using --im 'single' or 'inherited' [default = ENSG00000141510 (TP53 gene)]

	-f DATA_FOLDER, --data_folder=DATA_FOLDER
		Path to directory storing data files for simulated genome [default = ../data/genomes/hg38/]

	-o OUTPUT_FILE, --output_file=OUTPUT_FILE
		Path to output data file [default = ../data/output/test_out.rda]

	-h, --help
		Show this help message and exit

./src/run_v2_7.ipynb

• Jupyter notebook for running sompop_v0_2_7. Set up to perform single test runs and basic visualization of results, or batch runs.

./src/src_sompop_v0_2_7.r

• An R file containing function definitions for simulation

./src/src_sompop_DEV.r

• Source code for simulation incorporating a source of background mutations

./src/map_genome.ipynb

• Jupyter notebook for generating L1 target site annotations in the human genome

Data

./data/genomes

• Directory with sub-folders for each host genome that can be simulated. Contains only hg38 by default.
• Sub-folders hold data files required for simulation

./data/output

• Default folder for storing output data

./data/gene_lists

• Folder storing text files with lists of driver genes by ENSEMBL ID for simulating different tumor types

./data/genomes/hg38/drivers_XXXX.txt

• Text file containing a column of selected tumor suppressor genes for the tumor type XXXX (Lung squamous cell - LUSC, Glioblastoma multiforme - GBM, Colorectal adenocarcinoma - COREAD). Current driver gene lists were selected from IntOGen database: https://www.intogen.org/search

./data/genomes/hg38/drivers_XXXX_filt.txt

• Same as above except some filters have been applied to the lists

./data/genomes/hg38/exonicvsnon_counts.rda

• R data file containing the frequencies of L1 target sites of each snap-velcro category within exonic regions and outside exonic regions of the human genome. Exons were taken from Ensembl v94 human genome annotation.

./data/genomes/hg38/gene_pd_exon.rda

• R data file containing the probability of exonic L1 insertion in each protein-coding gene of the human genome based on weighted L1 target site count.

./data/genomes/hg38/gene_pdt_allsites.rda

• The same as gene_pdt.rda described above, though containing probabilities of insertion anywhere within the gene boundaries.