simulate_microbiome_counts: Simulate microbiome counts

View source: R/BRACoD.R

simulate_microbiome_countsR Documentation

Simulate microbiome counts

Description

Each bacteria's absolute abundance is simulated from a lognormal distribution. Then, convert each sample to relative abundance, and simulate sequencing counts using a multinomial distribution, based on the desired number of reads and the simulated relative abundances. This also simulates an environmental variable that is produced by some of the bacteria.

Usage

simulate_microbiome_counts(
  df,
  n_contributors = 20,
  coeff_contributor = 0,
  min_ab_contributor = -9,
  sd_Y = 1,
  n_reads = 1e+05,
  var_contributor = 5,
  use_uniform = TRUE,
  n_samples_use = NULL,
  corr_value = NULL,
  return_absolute = FALSE,
  seed = NULL
)

Arguments

df

A dataframe of OTU counts that is a model for data simulation. Samples are rows and bacteria are columns.

n_contributors

the number of bacteria that are to contribute to your environmental variable.

coeff_contributor

the average of the distribution used to simulate the contribution coefficient.

min_ab_contributor

The minimum log relative abundance, averaged across samples, to include a bacteria

sd_Y

the standard deviation of the simulated environmental variable

n_reads

the number of reads to be simulated per sample

var_contributor

If you use a uniform distribution, this is the range of the distribution, with a normal distribution it is the variance used to simulate the contribution coefficient.

use_uniform

use a uniform distribution to simulate the contribution coefficient. Alternative is the normal distribution.

n_samples_use

number of microbiome samples to simulate. If NULL, uses the same number of samples as in your dataframe

corr_value

the bacteria-bacteria correlation value you want to include in the simulation

return_absolute

returns the abosulte abundance values instead of the simulated microbiome counts

seed

random seed for reproducibility

Value

a list containing 1) the simulated count data 2) the simulated environmental variable and 3) the simulated contribution coefficients


BRACoD.R documentation built on March 24, 2022, 5:05 p.m.