This R package uses statistical modelling on data from NCBI databases and provides three statistical methods for genome size prediction of a given taxon, or group of taxa.
A straightforward weighted mean method (weighted-mean) identifies the closest taxa with available genome size information in the taxonomic tree and averages their genome sizes using weights based on taxonomic distance. A frequentist random effect model uses nested genus and family information to output genome size estimates. Finally, a third option provides predictions from a distributional Bayesian multilevel model which uses taxonomic information from genus all the way to superkingdom, therefore providing estimates and uncertainty bounds even for under-represented taxa.
All three methods use:
- A list of queries; a query being a taxon or a list of several taxa.
- A reference database containing all the known genome sizes, built from the NCBI databases, with associated taxa.
- A taxonomic tree structure as built by the NCBI.
genomesizeR retrieves the taxonomic classification of input queries, estimates the genome size of each query, and provides 95% confidence intervals for each estimate.
Install from GitHub:
install.packages("remotes")
remotes::install_github("ScionResearch/genomesizeR")
Download the archive containing the reference databases and the bayesian models from zenodo.org, using the inborutils package. You can change the path option to where you want to download the archive (default is current directory '.'):
remotes::install_github("inbo/inborutils")
inborutils::download_zenodo("10.5281/zenodo.13733183", path=".")
Store the path to the archive containing the reference databases and the bayesian models:
refdata_archive_path = "path/to/genomesizeRdata.tar.gz"
Read the example input file from the package:
example_input_file = system.file("extdata", "example_input.csv", package = "genomesizeR")
Load the package:
library(genomesizeR)
Run the main function to get the estimated genome sizes (with the default method which is the bayesian method):
results = estimate_genome_size(example_input_file, refdata_archive_path, sep='\t', match_column='TAXID', output_format='input')
Plot the genome size histogram per sample:
plotted_df = plot_genome_size_histogram(results)
Plot the genome size histogram for one sample:
plotted_df = plot_genome_size_histogram(results, only_sample='16S_1')
Plot the genome size boxplot per sample:
plotted_df = plot_genome_size_boxplot(results)
Plot the genome size boxplot for one sample:
plotted_df = plot_genome_size_boxplot(results, only_sample='ITS_1')
Plot the simplified taxonomic tree with colour-coded estimated genome sizes:
plotted_df = plot_genome_size_tree(results, refdata_archive_path)