The SYNY pipeline investigates gene collinearity (synteny) between genomes by reconstructing clusters from conserved pairs of protein-coding genes identified from DIAMOND homology searches. It also infers collinearity from pairwise genome alignments with minimap2 or MashMap3.
Show/hide: TOC
Show/hide: Introduction
Synteny is the measure of similarity in gene organization between two species. Through time, genomes can be reorganized by large-scale events such as recombination, but also smaller events like -- but not exhaustively -- duplication, translocation, deletion, or inversion.
Synteny inferences can be used to:
- Improve genome annotation (by helping to identify mispredicted and/or missing genes)
- Help differentiate between orthologous and paralogous genes (by leveraging positional information)
- Perform evolutionary distance analyses (e.g. species sharing similar/identical genome reorganization events are likely more closely related than species that do not share these events).
Show/hide: Installation
git clone https://github.com/PombertLab/SYNY.git
cd SYNY
export PATH=$PATH:$(pwd)SYNY dependencies can be installed automatically with Conda (does not require sudo privileges), with setup_syny.pl (requires sudo privileges), or manually.
Show/hide: Installing SYNY with Conda
SYNY can be installed without sudo-elevated privileges by leveraging conda packages. The installation process was tested with Miniconda3 on Ubuntu-22.04.3 LTS and Fedora 40 Linux distributions running as virtual machines on Microsoft Windows Subsystem for Linux (WSL).
To install SYNY with conda:
# Creating a conda environment
conda create -n syny
# Activating the conda environment
conda activate syny
## Installing syny within the conda environment
conda install syny -c conda-forge -c bioconda To run SYNY within its conda environment:
conda activate syny
(syny) username:~$ run_syny.pl -a *.gbff.gz -o output directoryIf a conda package manager is needed, Miniconda can be installed without the need for sudo-elevated privileges. To install Miniconda (tested on Ubuntu/Fedora):
## Setup variables
DOWN_DIR=$HOME/Downloads ## Replace by desired download directory
TOOLS_DIR=$HOME/Tools ## Replace by desired directory
CONDA_DIR=$TOOLS_DIR/miniconda3 ## Replace by desired subdirectory
CONFIG=~/.profile ## Desired configuration file
## Downloading miniconda
mkdir -p $DOWN_DIR $TOOLS_DIR
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh -P $DOWN_DIR
chmod +x $DOWN_DIR/Miniconda3-latest-Linux-x86_64.sh
## Installing miniconda
$DOWN_DIR/Miniconda3-latest-Linux-x86_64.sh \
-p $CONDA_DIR \
-b
echo "export PATH=\$PATH:$CONDA_DIR/condabin" >> $CONFIG
source $CONFIG
## Deleting downloaded file after installation
rm $CONDA_DIR/Miniconda3-latest-Linux-x86_64.sh
## Initializing conda
conda init
bash --login
## Updating conda and installing the libmamba solver
conda update --yes -n base conda
conda install --yes -n base conda-libmamba-solver
conda config --set solver libmamba
## Disabling automatic activation of the base environment at login
conda config --set auto_activate_base falseShow/hide: Installing SYNY with Perl install script
SYNY and its dependencies can be installed automatically with setup_syny.pl. This script will download and install DIAMOND, minimap2, MashMap3, Circos together with the required dnf/apt/zypper packages (this script has been tested on Fedora, Ubuntu, Debian, Kali, and openSUSE Tumbleweed distributions). Note that using this script will require sudo privileges to install dnf/apt/zypper packages.
## Listing supported Linux distributions:
setup_syny.pl --list_distro
Supported Linux distributions/package managers are:
debian apt
fedora dnf
kali apt
opensuse zypper
ubuntu apt## Installing dependencies:
LINUX=fedora ## Replace by Linux distribution: Ubuntu, Debian or Fedora
CONFIG=~/.bash_profile ## Replace by desired configuration file
DIR=~/TOOLS ## Replace by desired installation directory
setup_syny.pl \
--linux $LINUX \
--config $CONFIG \
--install $DIR
## Loading the configuration file:
source $CONFIGShow/hide: Installing SYNY manually
## On Ubuntu:
sudo apt install libperlio-gzip-perl
## On Fedora:
sudo dnf install perl-PerlIO-gzip## On Ubuntu:
sudo apt install python3-matplotlib
sudo apt install python3-seaborn
sudo apt install python3-pandas
sudo apt install python3-scipy
## On Fedora:
sudo dnf install python3-matplotlib
sudo dnf install python3-seaborn
sudo dnf install python3-pandas
sudo dnf install python3-scipy
## Or via pip (Ubuntu/Fedora):
pip install matplotlib
pip install seaborn
pip install pandas
pip install scipyversion=v2.1.9 ## Replace with desired DIAMOND version
DIR=/opt/diamond ## Replace with desired installation directory
mkdir -p $DIR
curl \
-L https://github.com/bbuchfink/diamond/releases/download/$version/diamond-linux64.tar.gz \
-o $DIR/diamond-linux64.tar.gz
tar -zxvf $DIR/diamond-linux64.tar.gz --directory $DIR
rm $DIR/diamond-linux64.tar.gz
export PATH=$PATH:$DIR## On Ubuntu
sudo apt install build-essential
sudo apt install zlib1g-dev
## On Fedora
sudo dnf group install "C Development Tools and Libraries"
sudo dnf install zlib-devel
## Ubuntu/Fedora
git clone https://github.com/lh3/minimap2
cd minimap2 && make
# To install the k8 javascript shell inside the minimap2 dir
curl \
-L https://github.com/attractivechaos/k8/releases/download/v0.2.4/k8-0.2.4.tar.bz2 \
-o k8-0.2.4.tar.bz2
tar -jxf k8-0.2.4.tar.bz2
cp k8-0.2.4/k8-`uname -s` k8
export PATH=$PATH:$(pwd) ## minimap2 install directory
export PATH=$PATH:$(pwd)/misc ## minimap2 subdir containing paftools.js## On Ubuntu
sudo apt install build-essential
sudo apt install gsl-bin
sudo apt install libgsl-dev
## On Fedora
sudo dnf group install "C Development Tools and Libraries"
sudo dnf install cmake
sudo dnf install gsl gsl-devel
git clone https://github.com/marbl/MashMap.git
cd MashMap
cmake -H. -Bbuild -DCMAKE_BUILD_TYPE=Release
cmake --build build
mv ./build/bin ./bin
export PATH=$PATH:$(pwd)/bin ## MashMap3 install directory## Dependencies (Ubuntu)
sudo apt install cpanminus
sudo apt install libgd-perl
## Dependencies (Fedora)
sudo dnf install perl-App-cpanminus
sudo dnf install perl-GD
## Dependencies (Ubuntu/Fedora)
sudo cpanm \
Text::Roman \
Clone \
Config::General \
Font::TTF::Font \
Getopt::ArgvFile \
List::MoreUtils \
Math::Bezier \
Math::Round \
Math::VecStat \
Params::Validate \
Readonly \
Regexp::Common \
Set::IntSpan \
Statistics::Basic \
SVG \
Text::Format
## Circos
wget \
--no-check-certificate \
https://circos.ca/distribution/circos-0.69-9.tgz
tar -zxvf circos-0.69-9.tgz
cd circos-0.69-9/bin
printf "\nexport PATH=\$PATH:$(pwd)" >> ~/.profile ## Ubuntu
printf "\nexport PATH=\$PATH:$(pwd)" >> ~/.bash_profile ## FedoraPerforming pairwise alignments between genomes can quickly consume a large amount of RAM. The amount of memory required depends on the size of the genomes being aligned and on the amounts of repetitive elements present in their sequences. When working with genomes larger than 150 Mbp, we recommend using MashMap3 instead of minimap2 as the genome alignment tool. While MashMap3 does not product exact alignments per se, its results are congruent with the exact alignments produced by minimap2 (and as such constitute good approximations). For example, alignments between two repeat-laden 500 Mbp genomes ran within 5 GB of RAM with MashMap3 whereas the same alignments with minimap2 peaked around 100 GB of RAM despite the two tools producing similar outcomes. However, note that when using MashMap3, reducing its minimum percentage identity threshold below its 85% default settings will significantly increase its RAM usage. If running above the maximum amount of memory available (real + virtual memory), pairwise genome alignments will be terminated by Linux out-of-memory (OOM) killer, resulting in blank outputs for the corresponding alignments and plots.
Show/hide: Using SYNY
The SYNY pipeline can be run with run_syny.pl, a master script that:
- Extracts genome/protein sequences and annotation data from GenBank (.gbf/.gbff) flat files.
- Performs round-robin pairwise genome alignments with minimap2 or MashMap3.
- Performs round-robin DIAMOND BLASTP homology searches, identifies conserved protein gene pairs, and reconstructs collinear clusters from these searches.
- Generates dotplots, barplots, linemaps and Circos plots highlighting collinear regions inferred from pairwise genome alignments and from shared protein cluster reconstructions.
- Generates heatmaps summarizing the percentages of collinear bases found in pairwise genome alignments and the percentages of proteins found in collinear clusters between all genomes.
When working with well-annotated, closely related genomes, collinear regions inferred from pairwise genome alignments and shared protein cluster reconstructions should yield similar results.
However, when working with genomes featuring high levels of sequence divergence, pairwise genome alignments may struggle. In those instances, collinear regions inferred from protein cluster reconstructions should outperform those from genome alignments; hypervariable intergenic regions are not considered in protein cluster reconstructions and silent mutations/codon usage biases do not affect amino acid sequences. Amino acids also have a larger character state and lower back-mutation probabilities than nucleotide sequences, allowing for searches across wider evolutionary distances.
Conversely, when working with poorly annotated or unannotated genomes, collinear regions inferred from genome alignments should outperform those inferred from protein cluster reconstructions (if there is sufficient sequence similarity to perform pairwise alignments).
SYNY can be run from the master script as follows:
run_SYNY.pl \
-a *.gbff \
-o SYNYThe run_syny.pl command line options for can be entered directly from the command line and/or provided from one or more configuration file(s) containing one entry per line (see commands.conf for an example). For example:
run_SYNY.pl \
-a *.gbff \
@commands.confOptions for run_SYNY.pl are:
-h (--help) Display all command line options
-t (--threads) Number of threads to use [Default: 16]
-p (--pthreads) Number of graphs to plot in parralel; defaults to --threads if unspecified
-a (--annot) GenBank GBF/GBFF Annotation files (GZIP files are supported)
-o (--outdir) Output directory [Default = SYNY]
-e (--evalue) DIAMOND BLASTP evalue cutoff [Default = 1e-10]
-g (--gaps) Allowable number of gaps between gene pairs [Default = 0]
--minsize Minimum contig size (in bp) [Default: 1]
--include Select contigs with names from input text file(s) (one name per line); i.e. exclude everything else
--ranges Select contigs with subranges from input text file(s): name start end
--exclude Exclude contigs with names matching the regular expression(s); e.g. --exclude '^AUX'
--aligner Specify genome alignment tool: minimap or mashmap [Default: minimap]
--asm Specify minimap max divergence preset (--asm 5, 10 or 20) [Default: off]
--mpid Specify mashmap percentage identity [Default: 85]
--resume Resume minimap/mashmap computations (skip completed alignments)
--no_map Skip minimap/mashmap pairwise genome alignments
--no_clus Skip gene cluster reconstructions
### Circos plots
-c (--circos) Circos plot mode: pair (pairwise), cat (concatenated), all (cat + pair) [Default: pair]
--orientation Karyotype orientation: normal, inverted or both [Default: normal]
--circos_prefix Prefix for concatenated plots [Default: circos]
-r (--ref) Reference to use for concatenated plots; uses first genome (alphabetically) if ommitted
-u (--unit) Size unit (Kb or Mb) [Default: Mb]
--winsize Sliding windows size (nucleotide biases) [Default: 10000]
--stepsize Sliding windows step (nucleotide biases) [Default: 5000]
--labels Contig label type: mixed (arabic + roman numbers), arabic, roman, or names [Default: mixed]
--label_size Contig label size [Default: 36]
--label_font Contig label font [Default: bold]
--custom_file Load custom colors from file
--list_preset List available custom color presets
--custom_preset Use a custom color preset, e.g.: --custom_preset chloropicon
--max_ticks Set max number of ticks [Default: 5000]
--max_ideograms Set max number of ideograms [Default: 200]
--max_links Set max number of links [Default: 75000]
--max_points_per_track Set max number of points per track [Default: 75000]
--clusters Color by cluster instead of contig/chromosome [Default: off]
--no_ntbiases Turn off nucleotide biases subplots
--no_cticks Turn off ticks in Circos plots
--no_circos Turn off Circos plots
### Barplots
-bh (--bheight) Barplot figure height in inches [Default: 10.8]
-bw (--bwidth) Barplot figure width in inches [Default: 19.2]
--bfsize Barplot font size [Default: 8]
--palette Barplot color palette [Default: Spectral]
--monobar Use a monochrome barplot color instead: e.g. --monobar blue
--bclusters Color clusters by alternating colors; colors are not related within/between contigs; [Default: off]
--bpmode Barplot mode: pair (pairwise), cat (concatenated), all (cat + pair) [Default: pair]
--no_barplot Turn off barplots
### Dotplots
-dh (--dheight) Dotplot figure height in inches [Default: 10.8]
-dw (--dwidth) Dotplot figure width in inches [Default: 19.2]
--dfsize Dotplot font size [Default: 8]
-m (--multi) Axes units multiplier (for dotplots) [Default: 1e5]
--color Dotplot color [Default: blue]
--dotpalette Use a color palette instead: e.g. --dotpalette inferno
--noticks Turn off ticks on x and y axes
--wdis Horizontal distance (width) between subplots [Default: 0.05]
--hdis Vertical distance (height) between subplots [Default: 0.1]
--no_dotplot Turn off dotplots
### Heatmaps
-hh (--hheight) Heatmap figure height in inches [Default: 10]
-hw (--hwidth) Heatmap figure width in inches [Default: 10]
--hfsize Heatmap font size [Default: 8]
--hmpalette Heatmap color palette [Default: winter_r]
--hmax Set maximum color bar value [Default: 100]
--hmin Set minimum color bar value [Default: 0]
--hauto Set color bar values automatically instead
--no_heatmap Turn off heatmaps
### Linear maps
-lh (--lheight) Linear map figure height in inches [Default: 5]
-lw (--lwidth) Heatmap figure width in inches [Default: 20]
--lm_rpalette Reference genome color palette [Default: tab20]
--lm_xpalette Target genome color palette [Default: Blues]
--lmrotation Contig name rotation [Default: 90]
--lfsize Font size [Default: 8]
--no_linemap Turn off linemaps
The output directory will be structured as follows:
ls -lah SYNY/
drwxr-xr-x 7 jpombert jpombert 4096 Apr 24 09:54 .
drwx------ 20 jpombert jpombert 4096 Apr 24 09:54 ..
drwxr-xr-x 6 jpombert jpombert 4096 Apr 24 09:54 ALIGNMENTS
drwxr-xr-x 6 jpombert jpombert 4096 Apr 24 09:55 CLUSTERS
drwxr-xr-x 2 jpombert jpombert 4096 Apr 24 09:54 LISTS
drwxr-xr-x 7 jpombert jpombert 4096 Apr 24 09:55 PLOTS
drwxr-xr-x 4 jpombert jpombert 4096 Apr 24 09:54 SEQUENCES
-rw-r--r-- 1 jpombert jpombert 526 Apr 24 09:55 error.log
-rw-r--r-- 1 jpombert jpombert 1574 Apr 24 09:57 syny.logGenome/protein sequences and gene lists extracted from GenBank GBFF files are located in the SEQUENCES/ and LISTS/ directories, respectively. Data files from minimap2 pairwise genome alignments and from gene cluster inferences are located in the ALIGNMENTS/ and CLUSTERS/ directories, respectively. Plots are located in PLOTS/.
The contents of the subdirectories are:
- ALIGNMENTS:
- PAF: pairwise genome alignments in the corresponding format (minimap2/MashMap3)
- MAF, ALN: pairwise genome alignments in the corresponding formats (minimap2)
- METRICS (minimap2):
- Alignment length vs. similarity scatter plots (in PNG format)
- Alignment metrics summaries (in plain TXT format)
- CLUSTERS:
- DIAMOND:
- DB
- BLASTP databases
- Round-robin BLASTP results (.diamond.6)
- DB
- HOMOLOGS:
- Lists of proteins that are conserved within the samples (.conserved)
- Homology and conservation summaries (.conserved_summary)
- Lists of unique proteins and their locations (.unique)
- Tab-delimited protein/description lists of proteins and their top homologs in other species (.shared.tsv)
- Tab-delimited protein/description lists of proteins uniques to each species (.uniques.tsv)
- Tab-delimited protein/description lists of all proteins, shared or not (.all.tsv)
- SYNTENY:
- Cluster summary (clusters_summary.tsv)
- Tab-delimited cluster summary table (clusters_summary_table.tsv)
- Subdirectory per specified gap allowance (gap_#) containing:
- CLUSTERS:
- Round-robin reconstructed syntenic clusters for each species
- PAIRS:
- Round-robin identified gene pairs for each species
- CLUSTERS:
- DIAMOND:
- LISTS:
- Lists of protein coding genes with location details (.list)
- PLOTS:
- BARPLOTS
- Barplots (in PNG/SVG format) from minimap2/mashmap3 alignments (.mmap.)
- Barplots (in PNG/SVG format) from protein clusters found with SYNY (.gap_0., .gap_1., ...)
- CIRCOS:
- Circos plots (in PNG/SVG format) from minimap2/mashmap3 alignments (.mmap.)
- Circos plots (in PNG/SVG format) from protein clusters found with SYNY (.gap_0., .gap_1., ...)
- CIRCOS_DATA:
- Configuration files for Circos plots
- DOTPLOTS:
- Dotplots (in PNG/SVG format) from minimap2/mashmap3 alignments (.mmap.)
- Dotplots (in PNG/SVG format) from protein clusters found with SYNY (.gap_0., .gap_1., ...)
- HEATMAPS:
- Heatmaps (in PNG/SVG format) summarizing the percentages of collinear bases in pairwise alignments (.mmap.)
- Heatmaps (in PNG/SVG format) summarizing the percentages of proteins found in clusters (.gap_0., .gap_1., ...)
- LINEMAPS:
- Linear maps (in PNG/SVG format) from minimap2/mashmap3 alignments (.mmap.)
- Linear maps (in PNG/SVG format) from protein clusters found with SYNY (.gap_0., .gap_1., ...)
- BARPLOTS
- SEQUENCES:
- GENOMES:
- FASTA files containing the sequences of the investigated genomes
- PROTEINS:
- Protein sequences for each species (.faa) in FASTA format
- GENOMES:
Below is a quick example describing how to compare a few select Cryptococcus genomes (C. neoformans var. neoformans strain JEC21, C. gattii strain WM276, C. gattii VGV strain MF34, C. decagattii strain 7685027, C. deuterogattii strain R265) using annotation data available from the public NCBI Genome database.
Downloading Cryptococcus example data automatically from NCBI using the provided Cryptococcus.sh shell script from the Examples/ subdirectory:
DATA=~/DATA ## Replace ~/DATA by desired annotation data directory
Cryptococcus.sh $DATAOr downloading the Cryptococcus data manually instead:
DATA=~/DATA ## Replace ~/DATA by desired annotation data directory
mkdir -p $DATA
##### Downloading data from NCBI ####
BASEURL=https://ftp.ncbi.nlm.nih.gov/genomes/all
### Cryptococcus neoformans strain JEC21
outfile=${DATA}/JEC21.gbff.gz
printf "\nDownloading Cryptococcus neoformans strain JEC21 as ${outfile}\n\n"
curl \
-L ${BASEURL}/GCF/000/091/045/GCF_000091045.1_ASM9104v1/GCF_000091045.1_ASM9104v1_genomic.gbff.gz \
-o $outfile
### Cryptococcus gattii strain WM276
outfile=${DATA}/WM276.gbff.gz
printf "\nDownloading Cryptococcus neoformans strain JEC21 as ${outfile}\n\n"
curl \
-L ${BASEURL}/GCF/000/185/945/GCF_000185945.1_ASM18594v1/GCF_000185945.1_ASM18594v1_genomic.gbff.gz \
-o $outfile
### Cryptococcus gattii VGV strain MF34
outfile=${DATA}/MF34.gbff.gz
printf "\nDownloading Cryptococcus gattii VGV strain MF34 as ${outfile}\n\n"
curl \
-L ${BASEURL}/GCA/009/650/685/GCA_009650685.1_Cryp_gatt_MF34/GCA_009650685.1_Cryp_gatt_MF34_genomic.gbff.gz \
-o $outfile
### Cryptococcus decagattii strain 7685027
outfile=${DATA}/D7685.gbff.gz
printf "\nDownloading Cryptococcus decagattii strain 7685027 as ${outfile}\n\n"
curl \
-L ${BASEURL}/GCA/036/417/295/GCA_036417295.1_ASM3641729v1/GCA_036417295.1_ASM3641729v1_genomic.gbff.gz \
-o $outfile
### Cryptococcus deuterogattii strain R265
outfile=${DATA}/R265.gbff.gz
printf "\nDownloading Cryptococcus deuterogattii strain R265 as ${outfile}\n\n"
curl \
-L ${BASEURL}/GCF/002/954/075/GCF_002954075.1_C._deuterogattii_R265_chr/GCF_002954075.1_C._deuterogattii_R265_chr_genomic.gbff.gz \
-o $outfileSYNY can be run from the command line with the run_syny.pl master script. In the command line below, no gap is allowed during gene cluster inferences, and Circos plots are produced in pairwise mode. The latter can be changed with the --circos command line switch (possible values are: pair, concatenated, all). Note that when comparing several genomes, concatenated plots can quickly become too dense for legibility. Producing concatenated plots can also significantly increase computation time.
##### Runtime (Intel i5-12500H mobile CPU)
# Total: 16 min; Circos: 11 min
DATA=~/DATA ## Replace by annotation data directory
SYNY=~/SYNY_RESULTS/CRYPT_ALL ## Replace by desired output directory
run_syny.pl \
--threads 16 \
--annot $DATA/*.gbff.gz \
--outdir $SYNYIn the command line below, SYNY is run on a subset of two genomes (JEC21 and WM276), this time allowing for different maximum gap thresholds between gene pairs.
##### Runtime (Intel i5-12500H mobile CPU)
# Total: 2 min; Circos: 1 min)
DATA=~/DATA ## Replace by annotation data directory
SYNY=~/SYNY_RESULTS/CRYPT_SUB ## Replace by desired output directory
run_syny.pl \
-t 16 \
-a $DATA/{JEC21,WM276}.gbff.gz \
-o $SYNY \
-g 0 1 5 \
-e 1e-10head -n 32 $SYNY/CLUSTERS/SYNTENY/clusters_summary.tsv
##### JEC21_vs_WM276; Gap = 0 #####
Total number of proteins in clusters: 5760
# of clusters: 779
Longest: 64
Shortest: 2
Average cluster size: 7
Median cluster size: 5
N50: 10
N75: 6
N90: 4
##### JEC21_vs_WM276; Gap = 1 #####
Total number of proteins in clusters: 5921
# of clusters: 228
Longest: 204
Shortest: 2
Average cluster size: 26
Median cluster size: 15
N50: 53
N75: 26
N90: 13
##### JEC21_vs_WM276; Gap = 5 #####
Total number of proteins in clusters: 5956
# of clusters: 53
Longest: 434
Shortest: 2
Average cluster size: 112
Median cluster size: 58
N50: 237
N75: 163
N90: 66head -n 22 $SYNY/CLUSTERS/SYNTENY/gap_0/CLUSTERS/JEC21_vs_WM276.clusters
### Cluster 0001; CNA00020 to CNA00070; CGB_B0020C to CGB_B0070W; size = 6 ###
CNA00020 - CGB_B0020C -
CNA00030 + CGB_B0030W +
CNA00040 - CGB_B0040C -
CNA00050 + CGB_B0050W +
CNA00060 - CGB_B0060C -
CNA00070 + CGB_B0070W +
### Cluster 0002; CNA00080 to CNA00150; CGB_B0080W to CGB_B0150C; size = 8 ###
CNA00080 + CGB_B0080W +
CNA00090 + CGB_B0090W +
CNA00100 - CGB_B0100C -
CNA00110 + CGB_B0110W +
CNA00120 - CGB_B0120C -
CNA00130 + CGB_B0130W +
CNA00140 - CGB_B0140C -
CNA00150 - CGB_B0150C -
### Cluster 0003; CNA00180 to CNA00220; CGB_B0180C to CGB_B0220W; size = 5 ###
CNA00180 - CGB_B0180C -
CNA00190 - CGB_B0190C -
CNA00200 + CGB_B0200W +
CNA00210 - CGB_B0210C -
CNA00220 + CGB_B0220W +Overall cluster metrics are summarized in CLUSTERS/SYNTENY/clusters_summary_table.tsv.
head -n 7 $SYNY/CLUSTERS/SYNTENY/clusters_summary_table.tsv
### Query Total # proteins Allowed Gaps Total # proteins in clusters % of proteins in clusters # of clusters Longest Shortest Average Median N50 N75 N90
JEC21_vs_WM276 6863 0 5762 83.96 779 64 2 7 5 10 6 4
JEC21_vs_WM276 6863 1 5922 86.29 228 204 2 26 15 53 26 13
JEC21_vs_WM276 6863 5 5957 86.80 53 434 2 112 58 237 163 66
WM276_vs_JEC21 6565 0 5760 87.74 778 64 2 7 5 10 6 4
WM276_vs_JEC21 6565 1 5923 90.22 229 204 2 26 15 53 26 13
WM276_vs_JEC21 6565 5 5957 90.74 51 477 2 117 56 260 164 66To facilitate comparisons when working with large datasets, heatmaps summarizing the percentages of collinear bases in pairwise genome alignments (.mmap.) and the percentages of protein coding-genes found in collinear clusters between each pair of genomes (e.g. .gap_0.) are generated with matplotlib.
In the above example, small heatmaps with 25 datapoints (5 * 5 genomes) will be generated for the minimap2 pairwise alignments and for the protein clusters found for each gap value investigated. In these heatmaps, percentages between pairs of genomes will vary based on the total number of bases and proteins found in the query used: i.e. (collinear bases / total bases in query) * 100 and (proteins in clusters / total proteins in query) * 100.
Example of a heatmap showing percentages of collinear protein-coding genes between investigated genomes:
Heatmap dimensions (default: 10 x 10) can be modified with the --hheight and --hwidth command line switches. Color bar min/max values (defaults: 0/100) can be changed with the --hmin and --hmax command line switches or calculated automatically with --hauto. The heatmap color palette (default: winter_r) can be changed with the --hmpalette command line switch (see this URL for a list of color palettes). Color palettes available on the operating system can be listed and/or plotted with check_mp_colors.py.
Unless the --no_circos command line switch is invoked, Circos (pairwise and/or concatenated) will be generated from the protein clusters identified with SYNY (e.g. .gap_0.) and/or from the genome alignments computed with minimap2 or MashMap3 (.mmap.).
In the pairwise plots (--circos pair), genomes are plotted in pairs (query vs. subject) using the query as the reference. In the concatenated plots (--circos cat), all genomes are plotted together in a single figure, using the reference genome specified with the --ref command line switch. If omitted, the first genome encountered alphabetically will be used as the default reference. Both concatenated and pairwise plots can be generated with the --circos all command line switch.
Karyotypes can be plotted with Circos in normal and/or inverted orientation(s). In the inverted plots, the contigs/chromosomes from the genome(s) being compared to the reference are plotted in reverse, from last to first. By default, the Circos plots are plotted in normal orientation. This behavior can be modified with the --orientation command line switch (possible values are normal, inverted and both).
In this figure, nucleotides biases are plotted in the concentric rings. These subplots can be skipped with the --no_ntbiases command line switch. From from outer to inner, the concentric rings are:
- AT and GC nucleotide biases (grey and red lines)
- GT and AC nucleotide biases (blue and green lines)
- GA and CT nucleotide biases (purple and yellow lines)
Syntenic blocks identified by SYNY are indicated by ribbons. These ribbons are color-coded based on the chromosomes/contigs present in the reference genome used. If desired, color coding can be set by cluster instead with --clusters. This option is useful when working with prokaryotes featuring a single chromosome, so that the ribbons are not all of the same color.
By default, contigs will be labelled by mixed numerals (roman + arabic) in the Circos plots. Contigs from the reference (query) will be labelled by roman numerals, others by arabic numerals. This behavior can be changed with the --labels command line option. Possible values are mixed, roman, arabic and names. Using --labels roman or --labels arabic will set all label numbers to the corresponding format, whereas using --labels names will label contigs by their names instead. Label sizes (default: 36) and fonts can be further adjusted with the --label_size and --label_font command line options. Possible fonts are: light,
normal, default, semibold, bold, italic, bolditalic, italicbold (see this Circos tutorial for details).
##### Runtime (Intel i5-12500H mobile CPU)
# Total: 2 min; Circos: 1 min
DATA=~/DATA ## Replace by annotation data directory
SYNY=~/SYNY_RESULTS/CRYPT_SUB_NAMES ## Replace by desired output directory
run_syny.pl \
-a $DATA/{JEC21,WM276}.gbff.gz \
-g 0 1 5 \
-e 1e-10 \
-r JEC21 \
-o $SYNY \
--circos all \
--circos_prefix WM276_vs_JEC21 \
--labels names \
--label_size 20 \
--label_font semibold
All data and configuration files for the Circos plots are in the PLOTS/CIRCOS_DATA/ subdirectory. If desired, chromosome labels can be further adjusted manually by editing the LABEL column in the corresponding Circos karyotype file(s). Karyotype files created by SYNY (in PLOTS/CIRCOS_DATA/ ) ends with *.karyotype.
## To see the first 5 lines from a karyotype file:
head -n 5 concatenated.normal.karyotype
#chr - ID LABEL START END COLOR
chr - NC_014938 1 0 1984822 black
chr - NC_014939 2 0 2187694 black
chr - NC_014940 3 0 1961511 black
chr - NC_014941 4 0 2233617 blackOnce edited, Circos plots can be regenerated by running Circos on the corresponding configuration files, e.g.:
SYNY=~/SYNY_RESULTS/CRYPT_SUB_NAMES ## Replace by SYNY results directory
OUTDIR=~/PLOTS ## Replace by desired output directory
circos \
-conf $SYNY/PLOTS/CIRCOS_DATA/concatenated/concatenated.gap_0.normal.conf \
-outputdir $OUTDIR \
-outputfile circos.gap_0.normal.pngExample of an image generated with Circos and SYNY from shared proteins clusters (using the inverted karyotype):
In the inverted karyotype image, the order of the karyotype(s) to be compared to the reference one is reversed. This option can be useful when comparing genomes whose chromosomes have been assigned similar numbers based on various inference methods (this does not appear to be the case in the above example). In such instances, inverting the karyotypes can help improve figure legibility.
In pairwise genome alignments, repetitive regions (such as telomeres/subtelomeres) can produce more than one alignment for a given locus. In the above figure, a bit of extra noise is added to the figure (as thin criss-crossing lines) due to these repetitive segments. As a rule of thumb, repetitive segments are easier to spot in dotplot-like figures (see Dotplots section below).
Chromosome maps (aka barplots) highlighting collinear genome segments will be generated with paf_to_barplot.py and matplotlib from minimap2/mashmap3 pairwise genome alignments (.mmap.) and from protein clusters identified with SYNY (e.g. .gap_0.).
In these plots, collinear regions found between the compared genomes are highlighted by colored rectangles. By default, these rectangles are color-coded based on the contigs/chromosomes of the query. The above barplot image was generated using the Spectral color palette from seaborn (set as default in SYNY). This palette can be replaced using the --palette command line switch; e.g. --palette husl (see this URL for a list of color palettes). Color palettes available on the operating system can be listed and/or plotted with check_mp_colors.py.
If desired, the barplots can instead be generated using a single monochromatic color with the --monobar command line switch; e.g. --monobar red. If this option is selected, the color-coding legend will be omitted from the plot.
By default, the barplots are formatted for a widescreen (landscape) output (width/height ratio: 19.2/10.8). This ratio can be adjusted with the --bheight and --bwidth command line switches.
By default, SYNY generates pairwise barplots. If desired, concatenated barplots can also be generated. In these concatenated barplots, the contigs from all species queried against a species (the reference) are plotted together. This behaviour can be changed with the --bpmode option. Possible values are pair (pairwise; set as default), cat (concatenated), and all (cat + pair).
Linear chromosome maps (aka linemaps) highlighting collinear genome segments will be generated with linear_maps.py and matplotlib from pairwise genome alignments (.mmap.) and from protein clusters identified with SYNY (e.g. .gap_0.). In these plots, collinear regions found between the compared genomes are highlighted by colored polygons. By default, these polygons are color-coded based on the contigs/chromosomes of the reference. The above linemap image was generated using the tab20/Blues color palettes for the reference/target genomes (both set as defaults in SYNY). These palettes can be replaced using the --lm_rpalette and --lm_xpalette command line switches; e.g. --lm_xpalette Grays (see URL for a list of color palettes). By default, the linemaps are formatted using a 20:5 width:height ratio. This ratio can be adjusted with the --lheight and --lwidth command line switches.
Unless the --no_dotplot command line switch is invoked, dotplots will be generated with paf_to_dotplot.py and matplotlib from pairwise genome alignments (.mmap.) and from protein clusters identified with SYNY, (e.g. .gap_0.).
In these dotplots, each chromosome/contig from the query is plotted as a column (x-axis) against each chromosome/contig from the subject (y-axis). In the above example, a total of 196 subplots (14 x 14 chromosomes) are plotted using matplotlib's subplot function. In these plots, matches are plotted using the selected color (defaut: blue), with repeated loci indicated by the presence of matches across two or more contigs/chromosomes.
Because large numbers quickly overlap in the small subplots, to improve legibility, numbers in the x and y axes are scaled down using a desired scale with the --multi command line switch (default: 1e5). For example:
1e3: n x 1000 bp
1e4: n x 10000 bp
1e5: n x 100000 bp
1e6: n x 1000000 bpAlternatively, ticks/numbers in the x and y axes can be turned off with the --noticks command line option.
By default, the dotplots are formatted for a widescreen (landscape) output (width/height ratio: 19.2/10.8). This ratio can be adjusted with the --dheight and --dwidth command line switches.
The default monochromatic color (blue) can be changed with the --color option (e.g. --color red). If desired, dotplots can instead be color-coded based on the query contigs/chromosomes with the --dotpalette option; e.g. --dotpalette husl (see this URL for a list of color palettes). Color palettes available on the operating system can be listed and/or plotted with check_mp_colors.py.
##### Runtime (Intel i5-12500H mobile CPU);
# Total: < 1 min; Circos: N/A
DATA=~/DATA ## Replace by annotation data directory
SYNY=~/SYNY_RESULTS/CRYPT_HUSL ## Replace by desired output directory
## Running SYNY with the husl color palette (for dotplots)
## and the --no_circos option to skip Circos plotting
run_syny.pl \
-a $DATA/{JEC21,WM276}.gbff.gz \
-g 0 1 5 \
-e 1e-10 \
-o $SYNY \
--dotpalette husl \
--no_circos
When computing pairwise genome alignments with minimap2, PAF metrics will also be calculated independently to help assess the outcome of these alignments. These metrics are available in the ALIGNMENTS/METRICS/ subdirectory, with alignment length vs. sequence similarity (%) scatter plots available in PNG format.
In the above plot, the average sequence identity for each alignment is calculated from the PAF files as follows:
Likewise, the total average sequence identity percentage listed in the PAF metrics insert is calculated by dividing the sum of all residue matches by the sum of all aligned block lengths.
As a rule of thumb, pairwise alignments featuring lower sequence identity percentages will produce fewer and/or more fragmented collinear segments.
PAF metrics cannot be calculated for MashMap3 aligments since the data required is missing from its PAF output files. MashMap3 does not calculate alignments explicitly (as per its intructional manual) but runs in a much smaller memory footprint than minimap2 when using its default percentage identity (--mpid 85). As such, it constitutes an interesting alternative when running into memory constraints. However, note that lowering the MashMap3 default threshold will significantly increase its memory usage.
Below is a quick example describing how to compare a few select genomes from the Encephalitozoonidae (Encephalitozoon/Ordospora species E. intestinalis strain ATCC 50506, E. hellem strain ATCC 50604, E. cuniculi strain ATCC 50602, O. colligata strain OC4 and O. pajunii strain FI-F-10) using annotation data from NCBI.
The Encephalitozoonidae genomes in this example are highly collinear yet display a high level of sequence divergence. As such, they constitute a good case scenario to assess the benefits of using gene clusters to complement inferences based on pairwise genome alignments.
Downloading genome data automatically from NCBI using the Encephalitozoonidae.sh shell script from the Examples/ subdirectory:
DATA=~/ENCE ## Replace ~/ENCE by desired annotation data directory
Encephalitozoonidae.sh $DATAOr downloading the Encephalitozoonidae data manually instead:
DATA=~/ENCE ## Replace ~/ENCE by desired annotation data directory
mkdir -p $DATA
##### Downloading data from NCBI ####
BASEURL=https://ftp.ncbi.nlm.nih.gov/genomes/all
## Encephalitozoon intestinalis ATCC 50506 telomere-to-telomere (T2T) genome
outfile=${DATA}/Ei50506.gbff.gz
printf "\nDownloading Encephalitozoon intestinalis ATCC 50506 as ${outfile}\n\n"
curl \
-L ${BASEURL}/GCA/024/399/295/GCA_024399295.1_ASM2439929v1/GCA_024399295.1_ASM2439929v1_genomic.gbff.gz \
-o $outfile
## Encephalitozoon hellem ATCC 50604 telomere-to-telomere (T2T) genome
outfile=${DATA}/Eh50604.gbff.gz
printf "\nDownloading Encephalitozoon hellem ATCC 50604 as ${outfile}\n\n"
curl \
-L ${BASEURL}/GCA/024/399/255/GCA_024399255.1_ASM2439925v1/GCA_024399255.1_ASM2439925v1_genomic.gbff.gz \
-o $outfile
## Encephalitozoon cuniculi ATCC 50602 telomere-to-telomere (T2T) genome
outfile=${DATA}/Ec50602.gbff.gz
printf "\nDownloading Encephalitozoon cuniculi ATCC 50602 as ${outfile}\n\n"
curl \
-L ${BASEURL}/GCA/027/571/585/GCA_027571585.1_ASM2757158v1/GCA_027571585.1_ASM2757158v1_genomic.gbff.gz \
-o $outfile
## Ordospora colligata OC4
outfile=${DATA}/OcOC4.gbff.gz
printf "\nDownloading Ordospora colligata OC4 as ${outfile}\n\n"
curl \
-L ${BASEURL}/GCF/000/803/265/GCF_000803265.1_ASM80326v1/GCF_000803265.1_ASM80326v1_genomic.gbff.gz \
-o $outfile
## Ordospora pajunii FI-F-10
outfile=${DATA}/OpFIF10.gbff.gz
printf "\nDownloading Ordospora pajunii FI-F-10 as ${outfile}\n\n"
curl \
-L ${BASEURL}/GCF/021/821/965/GCF_021821965.1_FI-F-10_v._1/GCF_021821965.1_FI-F-10_v._1_genomic.gbff.gz \
-o $outfileTo reduce runtime, we will skip the Circos plots with the --no_circos option.
##### Runtime (Intel i5-12500H mobile CPU);
# Total: 1 min; Circos: N/A
DATA=~/ENCE ## Replace by annotation data directory
SYNY=~/SYNY_RESULTS/ENCE ## Replace by desired output directory
run_syny.pl \
-a $DATA/*.gbff.gz \
-g 0 \
-e 1e-10 \
-o $SYNY \
--no_circosHere, while the pairwise alignments and gene cluster collinear inferences produce similar results within species (not shown), they differ substantially between species. For example, below are two barplots between Encephalitozoon hellem and Ordospora colligata. While the results are congruent, gene clusters inferences .gap_0. perform better than those based on pairwise alignments (.mmap.), the latter of which struggle due to the high levels of sequence divergence involved. These differences will also be reflected in the dotplots, heatmaps and Circos plots generated (not shown)
When working with genomes exhibiting a high level of divergence, both minimap2 and MashMap3 may fail to produce decent alignments. If that happens, we recommend using MashMap3 with a minimum percentage identity set lower than its default 85% threshold (e.g. --mpid 70). Note that doing so will increase its RAM usage and may introduce artefacts/noise in the plots produced.
Barplots (gene clusters):
Barplots (genome alignments; minimap2):
Barplots (genome alignments; mashmap; --mpid 85):
Barplots (genome alignments; mashmap; --mpid 70):
Because comparing more than two genomes can be useful, concatenated Circos plots can be generated by SYNY using the --circos all or the --circos cat command line switches. Below is a quick example of how to generate such plots using the genomes from the three Encephalitozoon species downloaded previously.
##### Runtime (Intel i5-12500H mobile CPU);
# Total: < 1 min; Circos: 17 sec
DATA=~/ENCE ## Replace by annotation data directory
SYNY=~/SYNY_RESULTS/SYNY_3_spp ## Replace by desired output directory
run_syny.pl \
-a $DATA/E*.gbff.gz \
-g 0 \
-e 1e-10 \
-r Ei50506 \
-o $SYNY \
--circos cat \
--circos_prefix encephalitozoon
In the above Circos plot, links between the reference and the queried genomes are color-coded based on the reference karyotype; links between the queried genomes are shown in light gray. In these plots, each chromosome/contig is plotted as a distinct ideogram.
Note that while SYNY can plot multiple genomes together with --circos cat / --circos all, adding too much data can quickly clutter the concatenated Circos plots (drawing too many ideograms can make the figure illegible). If the concatenated plots are too dense, drawing pairwise plots with --circos pair should help with readability. However, when comparing genomes featuring many chromosomes/contigs, cluttering can occur even in pairwise plots.
Custom colors for Circos plots can be loaded directly from tab-delimited text files containing color names and their associated RGB values (see custom_color_1.txt for an example). A few custom presets are also available to use in SYNY.
##### Runtime (Intel i5-12500H mobile CPU);
# Total: 1 min; Circos: 37 sec
DATA=~/ENCE ## Replace by annotation data directory
SYNY=~/SYNY_RESULTS/ENCE_CC ## Replace by desired output directory
COLORS=~/custom_color_2.txt ## Replace by desired custom color file
run_syny.pl \
-a $DATA/E*.gbff.gz \
-g 0 1 5 \
-e 1e-10 \
-r Ei50506 \
-o $SYNY \
--custom_file $COLORS \
--circos all \
--circos_prefix encephalitozoon_ccExample of an image generated with Circos and SYNY comparing a total of 3 genomes (using a custom color set):
run_syny.pl --list_preset
Available color presets:
blues 13 colors
chloropicon 20 colors
encephalitozoon 11 colors
##### Runtime (Intel i5-12500H mobile CPU);
# Total: 1 min; Circos: 37 sec
DATA=~/ENCE ## Replace by annotation data directory
SYNY=~/SYNY_RESULTS/ENCE_CC_PRESET ## Replace by desired output directory
run_syny.pl \
-a $DATA/E*.gbff.gz \
-g 0 1 5 \
-e 1e-10 \
-r Ei50506 \
-o $SYNY \
--circos all \
--circos_prefix encephalitozoon_blues \
--custom_preset blues
SYNY uses GenBank flat files (.gbff) containing DNA sequences and annotated features as input. However, when performing genome alignment-based comparisons, the latter annotated features are not required. Because users may want to compare genomes before their annotation becomes available (this process often comes in the late stages of genome sequencing projects), a simple FASTA to GBFF converter fasta_to_gbff.pl producing feature-less .gbff files is available in the Utils/ subdirectory.
To convert FASTA file(s) to feature-less GBFF files compressed in gzip format:
fasta_to_gbff.pl \
--fasta *.fasta.gz \
--outdir GBFF \
--gzipOptions for fasta_to_gbff.pl are:
-f (--fasta) FASTA file(s) to convert (gziped files are supported)
-o (--outdir) Output directory [Default: GBFF]
-g (--gzip) Compress the GBFF output files
-v (--verbose) Add verbosity
When compressing the GBFF output files, fasta_to_gbff.pl will use pigz if available, otherwise it will default to gzip.
Show/hide: Funding and acknowledgments
This work was supported in part by the National Institute of Allergy and Infectious Diseases of the National Institutes of Health (award number R15AI128627) to Jean-Francois Pombert. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Show/hide: How to cite
SYNY: a pipeline to investigate and visualize collinearity between genomes. Julian AT, Pombert JF. bioRxiv, 2024.05.09.593317. DOI: 10.1101/2024.05.09.593317.
If you use SYNY, please also cite the tool(s) used for genome alignments and/or protein sequence homology, as needed:
Minimap2: pairwise alignment for nucleotide sequences. Li H. Bioinformatics. 2018 Sep 15;34(18):3094-3100. doi: 10.1093/bioinformatics/bty191.
Minmers are a generalization of minimizers that enable unbiased local Jaccard estimation. Kille B, Garrison E, Treangen TJ, Phillippy AM. Bioinformatics. 2023 Sep 2;39(9):btad512. PMID: 37603771 PMCID: PMC10505501. doi: 10.1093/bioinformatics/btad512.
Sensitive protein alignments at tree-of-life scale using DIAMOND. Buchfink B, Reuter K, Drost HG. Nature Methods. 18, 366–368 (2021). doi: 10.1038/s41592-021-01101-x
Show/hide: References
Sensitive protein alignments at tree-of-life scale using DIAMOND. Buchfink B, Reuter K, Drost HG. Nature Methods. 18, 366–368 (2021). doi: 10.1038/s41592-021-01101-x
Minimap2: pairwise alignment for nucleotide sequences. Li H. Bioinformatics. 2018 Sep 15;34(18):3094-3100. doi: 10.1093/bioinformatics/bty191.
Minmers are a generalization of minimizers that enable unbiased local Jaccard estimation. Kille B, Garrison E, Treangen TJ, Phillippy AM. Bioinformatics. 2023 Sep 2;39(9):btad512. PMID: 37603771 PMCID: PMC10505501. doi: 10.1093/bioinformatics/btad512.
Circos: an information aesthetic for comparative genomics. Krzywinski M, Schein J, Birol I, Connors J, Gascoyne R, Horsman D, Jones SJ, Marra MA. Genome Res. 2009 Sep;19(9):1639-45. doi: 10.1101/gr.092759.109.
Matplotlib: A 2D graphics environment. Hunter, JD. Computing in Science & Engineering. 2007 9(3):90-95. doi: 10.1109/MCSE.2007.55.
Seaborn: statistical data visualization. Waskom, ML. Journal of Open Source Software. 2021 6(60): 3021. doi: 10.21105/joss.03021
Data structures for statistical computing in Python. McKinney W. Proceedings of the 9th Python in Science Conference. 2010:56-61. doi: 10.25080/Majora-92bf1922-00a.
SciPy 1.0: fundamental algorithms for scientific computing in Python.. Virtanen P, Gommers R, Oliphant TE, Haberland M, Reddy T, Cournapeau D, Burovski E, Peterson P, Weckesser W, Bright J, van der Walt SJ, Brett M, Wilson J, Millman KJ, Mayorov N, Nelson ARJ, Jones E, Kern R, Larson E, Carey CJ, Polat İ, Feng Y, Moore EW, VanderPlas J, Laxalde D, Perktold J, Cimrman R, Henriksen I, Quintero EA, Harris CR, Archibald AM, Ribeiro AH, Pedregosa F, van Mulbregt P; SciPy 1.0 Contributors. Nat Methods. 2020 Mar;17(3):261-272. doi: 10.1038/s41592-019-0686-2. Epub 2020 Feb 3. PMID: 32015543; PMCID: PMC7056644.
The genome of the basidiomycetous yeast and human pathogen Cryptococcus neoformans. Loftus BJ, Fung E, Roncaglia P, Rowley D, Amedeo P, Bruno D, Vamathevan J, Miranda M, Anderson IJ, Fraser JA, Allen JE, Bosdet IE, Brent MR, Chiu R, Doering TL, Donlin MJ, D'Souza CA, Fox DS, Grinberg V, Fu J, Fukushima M, Haas BJ, Huang JC, Janbon G, Jones SJ, Koo HL, Krzywinski MI, Kwon-Chung JK, Lengeler KB, Maiti R, Marra MA, Marra RE, Mathewson CA, Mitchell TG, Pertea M, Riggs FR, Salzberg SL, Schein JE, Shvartsbeyn A, Shin H, Shumway M, Specht CA, Suh BB, Tenney A, Utterback TR, Wickes BL, Wortman JR, Wye NH, Kronstad JW, Lodge JK, Heitman J, Davis RW, Fraser CM, Hyman RW. Science. 2005 Feb 25;307(5713):1321-4. doi: 10.1126/science.1103773. Epub 2005 Jan 13. PMID: 15653466; PMCID: PMC3520129.
Genome variation in Cryptococcus gattii, an emerging pathogen of immunocompetent hosts. D'Souza CA, Kronstad JW, Taylor G, Warren R, Yuen M, Hu G, Jung WH, Sham A, Kidd SE, Tangen K, Lee N, Zeilmaker T, Sawkins J, McVicker G, Shah S, Gnerre S, Griggs A, Zeng Q, Bartlett K, Li W, Wang X, Heitman J, Stajich JE, Fraser JA, Meyer W, Carter D, Schein J, Krzywinski M, Kwon-Chung KJ, Varma A, Wang J, Brunham R, Fyfe M, Ouellette BF, Siddiqui A, Marra M, Jones S, Holt R, Birren BW, Galagan JE, Cuomo CA. mBio. 2011 Feb 8;2(1):e00342-10. doi: 10.1128/mBio.00342-10. PMID: 21304167; PMCID: PMC3037005.
A New Lineage of Cryptococcus gattii (VGV) Discovered in the Central Zambezian Miombo Woodlands. Farrer RA, Chang M, Davis MJ, van Dorp L, Yang DH, Shea T, Sewell TR, Meyer W, Balloux F, Edwards HM, Chanda D, Kwenda G, Vanhove M, Chang YC, Cuomo CA, Fisher MC, Kwon-Chung KJ. mBio. 2019 Nov 12;10(6):e02306-19. doi: 10.1128/mBio.02306-19. PMID: 31719178; PMCID: PMC6851281.
Telomere-to-Telomere genome assemblies of human-infecting Encephalitozoon species. Mascarenhas Dos Santos AC, Julian AT, Liang P, Juárez O, Pombert JF. BMC Genomics. 2023 May 4;24(1):237. doi: 10.1186/s12864-023-09331-3. PMID: 37142951; PMCID: PMC10158259.
A new microsporidian parasite, Ordospora pajunii sp. nov (Ordosporidae), of Daphnia longispina highlights the value of genomic data for delineating species boundaries. de Albuquerque NRM, Haag KL, Fields PD, Cabalzar A, Ben-Ami F, Pombert JF, Ebert D. J Eukaryot Microbiol. 2022 May;69(3):e12902. doi: 10.1111/jeu.12902. Epub 2022 Mar 28. PMID: 35279911.
The Ordospora colligata genome: Evolution of extreme reduction in microsporidia and host-to-parasite horizontal gene transfer. Pombert JF, Haag KL, Beidas S, Ebert D, Keeling PJ. mBio. 2015 Jan 13;6(1):e02400-14. doi: 10.1128/mBio.02400-14. PMID: 25587016; PMCID: PMC4313915.