PoMeLo

PoMeLo is a novel R-based bioinformatics approach to identifying metabolic vulnerabilities of bacterial pathogens to host-directed therapeutics.

Running PoMeLo the first time, dependencies & R session info

To run PoMeLo, R & RStudio must be installed. PoMeLo was developed using R v4.2.2 with RStudio 2022.12.0.

When running PoMeLo the very first time, install all needed R packages by running the 6 lines at the start of the script (which are commented out). Note that if you install the required packages as suggested by RStudio, you will still need to manually run the last 4 lines of these 6:

pkgs = c("igraph","RColorBrewer", "hexbin", "scales","grid", "lattice", "gdata", "gridExtra", "ape","reshape2", "ggplot2", "seqinr", "phangorn", "fs", "hash","ggdendro", "phytools","openxlsx","coop","tidyverse", "rstudioapi", "aplot")
install.packages(pkgs)
install.packages("BiocManager", repos = "https://cloud.r-project.org")
library(BiocManager, warn.conflicts = FALSE)
BiocManager::install("remotes")
BiocManager::install("YuLab-SMU/treedataverse")

Attached base R packages: grid, stats, graphics, grDevices, utils, datasets, methods, base

Other attached packages: knitr_1.42, reticulate_1.28, rstudioapi_0.14, fs_1.6.1, lubridate_1.9.2, forcats_1.0.0, stringr_1.5.0, purrr_1.0.1, readr_2.1.4, tidyr_1.3.0, tibble_3.1.8, tidyverse_2.0.0, aplot_0.1.10, ggtreeExtra_1.8.1, ggtree_3.6.2, treeio_1.22.0, tidytree_0.4.2, dplyr_1.1.0, treedataverse_0.0.1, coop_0.6-3, openxlsx_4.2.5.2, phytools_1.5-1, maps_3.4.1, ggdendro_0.1.23, phangorn_2.11.1, seqinr_4.2-23, ggplot2_3.4.1, hash_2.2.6.2, reshape2_1.4.4, ape_5.7, gridExtra_2.3, gdata_2.18.0.1, lattice_0.20-45, scales_1.2.1, hexbin_1.28.2, RColorBrewer_1.1-3, igraph_1.4.0

Running PoMeLo

To run PoMeLo, clone this repo, open the PML_script.R file in RStudio, and then click "Source".

PoMeLo is closely linked to analyses using the BV-BRC website (https://www.bv-brc.org/), and requires files generated via BV-BRC. Here is a typical workflow incorporating both BV-BRC & PoMeLo:

Users will first need to use BV-BRC to select genomes of two related, preferably monophyletic, groups of bacterial species (target & non-target). If the user wishes to determine their groupings via analysis of genome loss via a phylogenetically independent contrast approach, see the PIC section below.
- When examining genomes in BV-BRC (under the 'genomes' tab) of interested taxa, we recommend including only genomes which are Complete (no WGS), and of Good Quality.
- When examining BV-BRC genomes we also recommend to add the "Species" column as one of the viewed columns when viewing and selecting genomes. This column commonly has updated taxonomic information and is superior to the name used in the initial "Genome Name" column.
- The genomes should be placed into separate BV-BRC Genome Groups. From the page listing the users' Genome Groups (https://www.bv-brc.org/workspace/yourusername@patricbrc.org/home/Genome%20Groups), next download both the target group & non-target group. Be sure to download each table as .csv and also be sure to not modify these .csv files after downloading, as Excel will drop zeros in genome IDs.
It is not required, but the last portion of the PoMeLo pipeline incorporates a BV-BRC phylogenetic tree. If the user would like to include a phylogeny of the two groups, first create a separate Genome Group in BV-BRC incorporating the species in your target and non-target groups - but for this Genome Group, be sure to only include one genome per "Species" column. The user will also need to save this Genome Group as a .csv file.
- Within BV-BRC, use the Bacterial Genome Tree tool (https://www.bv-brc.org/app/PhylogeneticTree) under the Tools & Services menu to create the phylogeny. Select your Genome Group with one genome per species, but otherwise use default settings (e.g. 100 genes) and have the output saved to your Genome Groups folder. After ~2-20 hours depending on your group size, the completed phylogeny will be saved to a subfolder.
- Find the folder with the completed phylogeny, and download the tree file ending in "_tree.nwk" - this file as well as the .csv of the Genome Group will be used in the phylogenetic portion of the pipeline.
With these BV-BRC files in hand, start the PoMeLo pipeline. Users will be prompted to select the required files. Note the file mapping_BVBRC_allECs.tab can be found in the /scripts subfolder.
All of the output files will be saved in your ~/code directory, with additional plots in the subfolders /supplemental_plots_ec_by_taxon_per_pathway & /supplemental_plots_taxon_by_pathway. The output files will have the taxon name and date appended.

Optional phylogenetic genome size analysis

To generate a phylogeny of bacterial species calculating genome size change via a PIC approach, we include a subpipeline: the PML_PICanalysis.Rmd file. This script incorporates both Python and R code, and will output a phylogeny as a .pdf file. This script requires the same BV-BRC phylogeny (.nwk) and associated table of genomes (.csv) as described earlier. To run this .Rmd pipeline, open the file in RStudio and select "Run" but click on "Run All".

Similar to running the main script the very first time, install all needed R & python packages by uncommenting and manually running just the lines at the start of the script (which are commented out). Next re-comment these lines and click on the "Run All" option.

Tips

For running the main PoMeLo pipeline, users can simply click "Source" and pop-up windows will appear when required input files need to be selected. The script will also provide notes as it progreses, indicating progress or warnings if there are errors. For example based on the input species names, the pipeline will select a taxonomic name that is included in all output filenames - the script informs the user of this name and suggests changing the filenames if it is incorrect. If the user does not wish to include a phylogeny in the final steps, they can simply select the 'cancel' button in the last two prompts for .nwk and .csv files, and the script will end there. Also note that if there are only two species in the analysis, the last steps incorporating a phylogeny will fail. When downloading genome lists from BV-BRC, note that the file will by default be called PATRIC_genome.csv. Change the name of your download to match which dataset it is.

When the pipeline finishes, there will be a list of 'warning messages' depending on your analysis. One common message looks like:

1: In download.file(listAurl[i], destinations[i], mode = "wb",  ... :
  cannot open URL 'ftp://ftp.bvbrc.org/genomes/2884482.5/2884482.5.PATRIC.pathway.tab': FTP status was '550 Requested action not taken; file unavailable'

This means that one or more genomes in your list were not downloaded because the pathway data is missing from BV-BRC.

Citation

If you use PoMeLo, please cite our publication:

PoMeLo: A systematic computational approach to predicting metabolic loss in pathogen genomes

Abigail Leigh Glascock, Eric Waltari, Gytis Dudas, Joan Wong, Vida Ahyong

bioRxiv 2023; https://www.biorxiv.org/search/Glascock%252C%252BDudas%252C%252BWong%252C%252BAhyong

The folders supplemental_materials_Treponema_run & supplemental_materials_Rickettsia_run contain the inputs and outputs for the Treponema example in Glascock et al & the Rickettsia example in the companion paper by Medicielo et al.:

Evolutionary genomics identifies host-directed therapeutics to treat intracellular bacterial infections

Josette Medicielo, Eric Waltari, Abigail Leigh Glascock, Gytis Dudas, Brian DeFelice, Ira Gray, Cristina M. Tato, Joan Wong, Vida Ahyong

bioRxiv 2023; https://www.biorxiv.org/search/Medicielo%252C%252BTato%252C%252BWong%252C%252BAhyong

czbiohub-sf / pomelo