RNAModR

RNAModR provides functions to map lists of genomic loci of RNA modifications to a reference mRNA transcriptome, and perform exploratory functional analyses of sites across the transcriptome trough visualisation and statistical analysis of the distribution of sites across transcriptome regions (5'UTR, CDS, 3'UTR).

RNAModR performs enrichment analyses to assess the statistical significance of the spatial and sequence-specific localisation of RNA modification sites relative to null sites. Null sites can be generated automatically or may be supplied manually in the form of a list of genomic loci.

Note that enrichment analyses results may depend critically on the choice and validity of null sites. For example, in establishing a list of null sites for 5-methylcytidine modifications, the non-uniformity in the actual distribution of cytosines across transcript sections makes a simple position permutation approach inappropriate for generating null sites.

---

This is production code! Use it at your own risk. That means that documentation may be incomplete and functions may return unexpected errors. It also means that functionality may change, and backward compatability with code based on older versions is not guaranteed. A full list of changes in the code base can be found in the NEWS file.

I appreciate and encourage any and all testing; for issues, please open an official Issue on the GitHub project site.

Table of contents

• Installing RNAModR
  • The GitHub way
• Getting started
• Downloadable transcriptome data
• Documentation
• FAQ
  • How do I interpret the p-values
• Contributors
• Licensing

---

Installing RNAModR

The GitHub way (requires the devtools package)

1. Make sure you the following R/Bioconductor packages are installed

   • AnnotationDbi
   • Biostrings
   • GenomeInfoDb
   • GenomicFeatures
   • GenomicRanges
   • magrittr
   • gplots
   • RSQLite
   • rtracklayer

   The recommended way to install R/Bioconductor packages is to use BiocManager::install:

   BiocManager::install(c("AnnotationDbi", "Biostrings", "GenomeInfoDb", "GenomicFeatures", "GenomicRanges", "gplots", "RSQLite", "rtracklayer"))
   

   Additionally, RNAModR requires two organism-specific R packages to contruct a custom transcriptome. Currently, RNAModR supports human and mouse data, based on the following reference genome versions

   • Human: hg38, hg19
   • Mouse: mm10, mm9.

   Please install the corresponding organism- and version-matching R/Bioconductor packages. For example, if genomic loci of RNA modification are based on the human GRCh38/hg38 reference genome, RNAModR requires the following R/Bioconductor packages:

   • BSgenome.Hsapiens.UCSC.hg38
   • org.Hs.eg.db

   which you can install in the usual way

   BiocManager::install(c("BSgenome.Hsapiens.UCSC.hg38", "org.Hs.eg.db"))
   

   We also offer the possibility to download pre-constructed transcriptome data, see section Downloadable transcriptome data.

2. If all package dependencies are met, install RNAModR with devtools

   if (!require("devtools")) install.packages("devtools")
   devtools::install_github("mevers/RNAModR", build_vignettes = FALSE)
   

   You can force a re-install of RNAModR by adding

   devtools::install_github(..., force = TRUE)
   

Getting started

The following lines of R code will load the RNAModR library, and plot the distribution of m6A sites [Linder et al., Nature Methods 12, 767 (2015)] across the 5'UTR, CDS and 3'UTR of the human hg38-based transcriptome.

Note: You can also download pre-constructed transcriptome data, see the next section for details.

# Load the library.
library(RNAModR)
library(magrittr)

# Build reference transcriptome.
# This might take a few minutes.
BuildTx("hg38")

# Load and map m6A sites to reference transcriptome.
m6A <- system.file("extdata", "miCLIP_m6A_Linder2015_hg38.bed", package = "RNAModR") %>%
    ReadBED() %>%
    SmartMap("m6A_Linder")

# Keep sites located in the 5'UTR, CDS and 3'UTR
m6A <- m6A %>%
    FilterTxLoc(c("5'UTR", "CDS", "3'UTR"))

# Generate null sites
null <- m6A %>%
    GenerateNull(nt = "A")

# Plot distribution across transcript sections
PlotSpatialEnrichment(m6A, null)

Downloadable transcriptome data

It is recommended to always custom-build transcriptome data using BuildTx().

Version	Organism	Assembly link
0.2.2	Homo sapiens	hg38
0.2.2	Homo sapiens	hg19
0.2.2	Homo sapiens	hg18
0.2.2	Mus musculus	mm10
0.2.2	Mus musculus	mm9
0.2.2	Mus musculus	mm8
<=0.1.1	Homo sapiens	hg38
<=0.1.1	Homo sapiens	hg19
<=0.1.1	Homo sapiens	hg18
<=0.1.1	Mus musculus	mm10
<=0.1.1	Mus musculus	mm9
<=0.1.1	Mus musculus	mm8

---

In order to use the transcriptome data you need to copy the RData file into your working directory. You can check that RNAModR correctly finds the transcriptome data by running e.g.

BuildTx("hg38")

Provided you have copied the file tx_hg38.RData into the working directory, this should produce the following message

Found existing transcriptome data. Nothing to do.
To rebuild run with force = TRUE.

Documentation

The most current RNAModR manual can be downloaded here.

FAQ

How do I interpret the p-values?

A man who ‘rejects’ a hypothesis provisionally, as a matter of habitual practice, when the significance is at the 1% level or higher, will certainly be mistaken in not more than 1% of such decisions. For when the hypothesis is correct he will be mistaken in just 1% of these cases, and when it is incorrect he will never be mistaken in rejection. [...] However, the calculation is absurdly academic, for in fact no scientific worker has a fixed level of significance at which from year to year, and in all circumstances, he rejects hypotheses; he rather gives his mind to each particular case in the light of his evidence and his ideas.

-- Sir Ronald A. Fisher, from Statistical Methods and Scientific Inference (1956)

Some advice concerning interpreting p-values

1. Be critical and conservative when interpreting statistical significance from the enrichment/depletion analyses.
2. Pay attention to the effect size (e.g. the odds-ratio).
3. p-values are corrected for multiple hypothesis testing assuming individual hypothesis tests (and their associated p-values) to be independent; this may not necessarily be justified. To err on the side of caution and to be conservative, p-values are adjusted using the Bonferroni correction.

Contributors

Please contact Maurits Evers in case of questions/suggestions. In case of bugs/feature requests please open an issue on github.

Licensing

The RNAModR R package is open source licensed under the GNU Public License, version 3 (GPLv3).