redst4r / birthdeathpy

Discrete Self Replicating Birth Death Process

Discrete Self Replicating Birth Death Process

Simulates a simple. discrete time stochastic model of a self replicating object, i.e.

A -> A + A
A -> 0

In addition, objects can mutate, i.e. A->B, and B replicates with the same rules.

Examples:

• cellular growth and death (mutation rate would e.g. be the emergence of a new genotype)
• PCR amplification

Installation

Requirements: Just a recent python3 installation with pip, e.g. miniconda.

pip install git+https://github.com/redst4r/birthdeathpy.git

Usage

Via the script birthdeath:

birthdeath simulate_uniform_abundance ...

This simulates a population of species with equal initial abundance.

birthdeath simulate_dirichlet_abundance ...

Simulates the population of species with initial abundances drawn from a Dirichlet distribution (--alpha parameter!). This allows for skewed starting distributions.

Example

birthdeath simulate_uniform_abundance \
  --n_species 1000 \
  --abundance 1 \
  --rounds 20 \
  --mutation_rate 0.001 \
  --death_probability 0.2 \
  -output_prefix results_

This initializes the population with 1000 distinct species, each having one individual (A_i = 1, i=1...1000), replicates for 20 rounds with given death and mutation probabilities (per round) and saves the results under results_....

Outputs

Outputs are saved under the given --output_prefix:

• ...clonesizes.csv: Sizes of all surviving species after replication
• ...initial_clonesize.png: Histogram of initial clonesizes
• ...final_clonesize.png: Histogram of final clonesizes