---
title: "miaSim: Microbiome Data Simulation"
author:
- name: Daniel Rios Garza
email: danielrios.garza@kuleuven.be
- name: Emma Gheysen
email: emma.gheysen@student.kuleuven.be
- name: Karoline Faust
email: karoline.faust@kuleuven.be
- name: Leo Lahti
email: leo.lahti@iki.fi
- name: Yagmur Simsek
email: yagmur.simsek@hsrw.org
- name: Yu Gao
email: gaoyu19920914@gmail.com
date: "`r Sys.Date()`"
package: miaSim
output:
BiocStyle::html_document:
fig_height: 7
fig_width: 10
toc: yes
toc_depth: 2
number_sections: true
vignette: >
%\VignetteIndexEntry{miaSim}
%\VignetteEngine{knitr::rmarkdown}
%\VignetteEncoding{UTF-8}
\usepackage[utf8]{inputenc}
---
```{r, echo=FALSE}
knitr::opts_chunk$set(cache = FALSE,
fig.width = 9,
message = FALSE,
warning = FALSE)
```
# Introduction
`miaSim` implements tools for microbiome data simulation based on the
`SummarizedExperiment` [@SE], `microsim` .
Microbiome time series simulation can be obtained by generalized
Lotka-Volterra model,`simulateGLV`, and Self-Organized Instability
(SOI), `simulateSOI`. Hubbell's Neutral model, `simulateHubbell` is used
to determine the species abundance matrix. The resulting abundance matrix
from these three simulation models is applied to `SummarizedExperiment`
object or `TreeSummarizedExperiment` object.
`powerlawA` and `randomA` give interaction matrix of species
generated by normal distribution and uniform distribution, respectively.
These matrices can be used in the simulation model examples.
`tDyn` generates lists of time series that can be specified as simulation time
and time points to keep in simulated time.
# Installation
### Bioc-release
```{r install-bioc,eval=FALSE}
if (!requireNamespace("BiocManager", quietly = TRUE))
install.packages("BiocManager")
```
# Models for simulating microbiome data sets
`simulateGLV` is the generalized Lotka-Volterra simulation model fitted to
time-series estimates microbial population dynamics and relative rates of
interaction. The model relies on interaction matrix that represents interaction
heterogeneity between species. This interaction matrix can be generated
with `powerlawA` or `randomA` functions depending on the distribution method.
`powerlawA` uses normal distribution to create interaction matrix.
```{r}
library(miaSim)
A_normal <- powerlawA(n.species = 4, alpha = 3)
```
`randomA` uses uniform distribution to create interaction matrix.
```{r}
A_uniform <- randomA(n.species = 10, d = -0.4, min.strength = -0.8,
max.strength = 0.8, connectance = 0.5)
```
The number of species specified in the interaction matrix must be the same
amount as the species used in the `simulateGLV` and `simulateSOI` models.
```{r}
SEobject <- simulateGLV(n.species = 4, A_normal, t.end = 1000)
```
Time series is added to `simulateGLV` with `tDyn` function where the time
points can be kept and extracted from simulation time as a separate list.
```{r}
Time <- tDyn(t.start = 0, t.end = 100, t.step = 0.5, t.store = 100)
Time$t.index
```
`simulateHubbell` includes the Hubbell Neutral simulation model which explains
the diversity and relative abundance of species in ecological communities.
This model is based on the community dynamics; migration, births and deaths.
```{r}
A_uniform <- randomA(n.species = 10, d = -0.4, min.strength = -0.8,
max.strength = 0.8, connectance = 0.5)
```
The number of species specified in the interaction matrix must be the same
amount as the species used in the `simulateGLV` and `simulateSOI` models.
```{r}
SEobject <- simulateGLV(n.species = 4, A_normal, t.start = 0, t.store = 1000)
```
Time series is added to `simulateGLV` with `tDyn` function where the time
points can be kept and extracted from simulation time as a separate list.
```{r}
Time <- tDyn(t.start = 0, t.end = 100, t.step = 0.5, t.store = 100)
Time$t.index
```
`simulateHubbell` includes the Hubbell Neutral simulation model which explains
the diversity and relative abundance of species in ecological communities.
This model is based on the community dynamics; migration, births and deaths.
```{r}
ExampleHubbell <- simulateHubbell(n.species = 8, M = 10, I = 1000, d = 50,
m = 0.02, tend = 100)
```
The Self-Organised Instability (SOI) model can be found in `simulateSOI` and it
generates time series for communities and accelerates stochastic simulation.
```{r}
ExampleSOI <- simulateSOI(n.species = 4, I = 1000, A_normal, k=5, com = NULL,
tend = 150, norm = TRUE)
```
The simulations result in the `SummarizedExperiment` [@SE] class object
containing the abundance matrix. Other fields, such as rowData containing
information about the samples, and colData, consisting of sample metadata
describing the samples, can be added to the `SummarizedExperiment` class object.
# Session info
```{r}
sessionInfo()
```