CompHeatmaps

Description

A large problem in conducting metatranscriptomic analysis on microbial communities is dealing with the complexity of the large amount of data researchers tend to work with. Heatmaps are a useful visualization that illustrate the abundance across these samples at a glance. However, going from raw sequence reads to heatmaps is quite a cumbersome process.

The goal of CompHeatmaps is to simplify this complex process by easily generating heatmaps from raw 16S sequence variants.

This package was created in R version 4.3.1 on Mac platform with macOS version Sonoma 14.1

Installation

You can install the development version of CompHeatmaps from GitHub with:

install.packages("devtools")
library("devtools")
devtools::install_github("kojiwong/CompHeatmaps", build_vignettes = TRUE)
library("CompHeatmaps")

To run the Shiny app:

CompHeatmaps::runCompHeatmaps()

Overview

The main components of this R package include the functions preprocess_16s_data, create_abundance_table, and create_heatmap. These functions are meant to be run in the specified listed order.

The image below illustrates what the inputs and the outputs of the package functions are at a glance.

The heatmap below is produced after create_heatmap() following the examples.

ls("package:CompHeatmaps")
browseVignettes("CompHeatmaps")

CompHeatmaps contains 3 functions: - preprocess_16s_data: This function takes in a directory of raw sequence reads of a sample in fastq file, one file per sample and an output directory where filtered results will be outputted to. In addition, it can take a verbose bool to print updates of preprocessing and a multithread bool if you wish to use multithreading. This function returns a list of a filtered dada objects which we can pass into create_abundance_table.

• create_abundance_table: This function takes in a list of dada objects to create an abundance table from preprocessed 16S data. You may also pass a multithread bool to process large amounts of data quicker. This function outputs a normalized and scaled matrix of relative abundance of sequence variants between samples.

• create_heatmap: This function takes in two inputs. The first is the abundance matrix we generated in create_abundance_table that we wish to visualize. The second parameter is proportion, a number between 0 and 1 which dictates the proportion of the abundance table we wish to visualize. By default, proportion is set to 1 which will visualize relative abundance of all the sequence variants. Ths function returns a ggplot2 heatmap.

The package also contains some sample data acquired from the Human Microbiome Project (HMP). This data is located in the directory “inst/extdata/sample_raw_16S_data” of this package.

Contributions

The author of this package is Koji Wong. The author wrote the preprocess_16s_data and create_abundance_table functions which utilizes the dada2 package to preprocess 16S rRNA data and formats it in an abundance table which can be utilized for the other function this author created create_heatmap, which calls the ggplot2 package for visualization.

This package heavily depends on the metatranscriptomic pipeline dada2 for data preprocessing and creation of abundance matrices. This package also utilizes ggplot2 to create heatmap illustrations of sequence variant abundances.

This package also makes use of ggplot2 to create heatmaps, as well as packages tibble and tidyr to clean up the table for ggplot2 input.

References

• A framework for human microbiome research. Human Microbiome Project Consortium, Nature, 486 (2012), pp. 215-221.

• Structure, function and diversity of the healthy human microbiome. Human Microbiome Project Consortium, Nature, 486 (2012), pp. 207–214.

• Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP (2016). “DADA2: High-resolution sample inference from Illumina amplicon data.” Nature Methods, 13, 581-583. doi:10.1038/nmeth.3869 https://doi.org/10.1038/nmeth.3869.

• Alexander Lex, Nils Gehlenborg, Hendrik Strobelt, Romain Vuillemot, Hanspeter Pfister, UpSet: Visualization of Intersecting Sets, IEEE Transactions on Visualization and Computer Graphics (InfoVis ’14), vol. 20, no. 12, pp. 1983–1992, 2014.

• Wickham H, Hester J, Chang W, Bryan J (2022). devtools: Tools to Make Developing R Packages Easier. R package version 2.4.5, https://CRAN.R-project.org/package=devtools.

• H. Wickham. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York, 2016.

• Müller K, Wickham H (2023). tibble: Simple Data Frames. R package version 3.2.1, https://CRAN.R-project.org/package=tibble.

• Wickham H, Vaughan D, Girlich M (2023). tidyr: Tidy Messy Data. R package version 1.3.0, https://CRAN.R-project.org/package=tidyr.

• Chang W, Cheng J, Allaire J, Sievert C, Schloerke B, Xie Y, Allen J, McPherson J, Dipert A, Borges B (2023). shiny: Web Application Framework for R. R package version 1.8.0, https://CRAN.R-project.org/package=shiny.

Acknowledgements

This package was developed as part of an assessment for the BCB410H class taken in Fall 2023 with Professor Anjali Silva at the University of Toronto, Toronto, Canada. CompHeatmaps welcomes any suggestions, issues, and contributions to this package.

Future Improvements

Currently this package produces a heatmap illustrating abundance of sequence reads across the samples after filtering, which is not so interesting as opposed to taxons. As a result, much of the heatmap will consist of neutral abundance representation.

To improve on this for the future, one goal of the package can be to perform taxonomic classification on the sequence table and some how incorporate this into heatmap creation rather than sequence variants.

Another improvement could be on the heatmap generation itself, by adding placeholder names for the sequence variants rather than the DNA sequence.