Multi-region sampling • tumopp

Workflow

Perform simulation and extract population data:

library(tumopp)
library(dplyr)
result = tumopp::tumopp("-N20000 -D3 -Chex -Lconst -k10")
population = result$population[[1L]]
graph = result$graph[[1L]]

Sample cells and put neutral mutations on the lineages:

extant = population |> tumopp::filter_extant()
ncell = 200L
regions = tumopp::sample_uniform_regions(extant, nsam = 4L, ncell = ncell)
subgraph = tumopp::subtree(graph, unlist(regions$id))
vaf = tumopp::make_vaf(subgraph, regions$id, mu = 8.0) |> print()
#> # A tibble: 29,816 × 4
#>      `1`   `2`   `3`   `4`
#>    <dbl> <dbl> <dbl> <dbl>
#>  1     0     0     0 0.005
#>  2     0     0     0 0.005
#>  3     0     0     0 0.005
#>  4     0     0     0 0.005
#>  5     0     0     0 0.005
#>  6     0     0     0 0.005
#>  7     0     0     0 0.005
#>  8     0     0     0 0.005
#>  9     0     0     0 0.025
#> 10     0     0     0 0.025
#> # ℹ 29,806 more rows

Estimate \(F_{ST}\) from VAF.

tumopp::dist_vaf(vaf, ncell) |> tumopp::fst()
#> [1] 0.06250332

# True FST from cell genealogy
tumopp::dist_genealogy(subgraph, regions$id) |> tumopp::fst()
#> [1] 0.06050219

Summarize and visualize VAF:

vaf_tidy = vaf |>
  tumopp::filter_detectable(0.01) |>
  tumopp::sort_vaf() |>
  tumopp::longer_vaf() |>
  print()

library(ggplot2)
ggplot(vaf_tidy) +
  aes(sample, site) +
  geom_tile(aes(fill = frequency)) +
  scale_fill_distiller(palette = "Spectral", limit = c(0, 1), guide = FALSE) +
  coord_cartesian(expand = FALSE)