mcmc

Function to sample the joint posterior distribution for the fitness value of all mutant and neutral linages given a time-series barcode count.

This function expects the data in a tidy format. This means that every row represents a single observation. For example, if we measure barcode i in 4 different time points, each of these four measurements gets an individual row. Furthermore, measurements of barcode j over time also get their own individual rows.

The DataFrame must contain at least the following columns:

• id_col: Column identifying the ID of the barcode. This can the barcode sequence, for example.
• time_col: Column defining the measurement time point.
• count_col: Column with the raw barcode count.
• neutral_col: Column indicating whether the barcode is from a neutral lineage

or not.

Keyword Arguments

• data::DataFrames.AbstractDataFrame: Tidy dataframe with the data to be

used to sample from the population mean fitness posterior distribution.

• n_walkers::Int: Number of walkers (chains) for the MCMC sample.
• n_steps::Int: Number of steps to take.
• outputname::String: String to be used to name the .jld2 output file.
• model::Function: Turing.jl model defining the posterior distribution from which to sample (see BarBay.model module). This function must take as the first four inputs the following:
  • R̲̲::Array{Int64}:: 2 or 3D array containing the raw barcode counts for all tracked genotypes. The dimensions of this array represent:
    • dim=1: time.
    • dim=2: genotype.
    • dim=3 (optional): experimental repeats
  • n̲ₜ::VecOrMat{Int64}: Array with the total number of barcode counts for each time point (on each experimental repeat, if necessary).
  • n_neutral::Int: Number of neutral lineages.
  • n_bc::Int: Number of neutral lineages.

Optional Keyword Arguments

• model_kwargs::Dict=Dict(): Extra keyword arguments to be passed to the model function.
• id_col::Symbol=:barcode: Name of the column in data containing the barcode identifier. The column may contain any type of entry.
• time_col::Symbol=:time: Name of the column in data defining the time point at which measurements were done. The column may contain any type of entry as long as sort will resulted in time-ordered names.
• count_col::Symbol=:count: Name of the column in data containing the raw barcode count. The column must contain entries of type Int64.
• neutral_col::Symbol=:neutral: Name of the column in data defining whether the barcode belongs to a neutral lineage or not. The column must contain entries of type Bool.
• rep_col::Union{Nothing,Symbol}=nothing: Optional column in tidy dataframe to specify the experimental repeat for each observation.
• rm_T0::Bool=false: Optional argument to remove the first time point from the inference. Commonly, the data from this first time point is of much lower quality. Therefore, removing this first time point might result in a better inference.
• sampler::Turing.Inference.InferenceAlgorithm=Turing.NUTS(0.65): MCMC sampler to be used.
• ensemble::Turing.AbstractMCMC.AbstractMCMCEnsemble=Turing.MCMCSerial():

Sampling modality to be used. Options are: - Turing.MCMCSerial() - Turing.MCMCThreads() - Turing.MCMCDistributed()

• verbose::Bool=true: Boolean indicating if the function should print partial progress to the screen or not.