Version: 2
Authors and Contributors: Catherine Mageeney, Corey Hudson, Steven Yu, Ellis Torrance, Kelly Williams
If you're using this software in a publication, please cite:
- Mageeney CM, Trubl G, Williams KP. 2022. Improved mobilome delineation in fragmented genomes. Front Bioinform doi: 10.3389/fbinf.2022.866850
- Yu SL, Mageeney CM, Shormin F, Ghaffari N, Williams KP. Speeding genomic island discovery through systematic design of reference database composition. PLoS One. 2024 Mar 13;19(3):e0298641. doi: 10.1371/journal.pone.0298641.
- Mageeney CM, Lau BY, Wagner JW, Hudson CM, Schoeniger JS, Krishnakumar R, Williams KP. 2020. New candidates for regulated gene integrity revealed through precise mapping of integrative genetic elements. Nucleic Acids Research 48(8):4052-4065 (doi: 10.1093/nar/gkaa156)
- Hudson CM, Lau BY, Williams KP. 2015. Islander: a database of precisely mapped genomic islands in tRNA and tmRNA genes. Nucleic Acids Research V43(D1):D48–D53 (doi: 10.1093/nar/gku1072)
Within this repository are two main programs: Islander and TIGER. These two programs use different methodology for identifying putative integrative genetic elements in genomic assemblies.
To identify genomic islands the Islander software looks for tyrosine integrase genes located within either a tmRNA or tRNA gene. It first identifies tRNA and tmRNA genes using tRNAscan-SE and ARAGORN. It then identifies nearby integrases (excluding Xer and integron subclasses) using an integrase specific HMM with HMMER3. Each identified tRNA or tmRNA then serves as BLASTN query in a search for the t(m)RNA gene fragment that marks the other end of the putative integrated genomic island. Each candidate island is then subjected to a series of tests to determine whether they represent true integrative genomic islands (see image below).
Graphical Description of the Islander algorithm. (A and B) Population Phase: tRNA and tmRNA genes (tDNAs), tDNA fragments and integrase genes (ints) are mapped on the chromosome, and each int-bearing interval between a tDNA and its cognate fragments is considered a candidate island. (C) Filtering Phase: Candidates pass through several filters, including tests for an integrase gene, correct fragment/tDNA orientation and length. (D) Resolution Phase: Multiple candidates at the same tDNA are resolved, disallowing overlaps and identifying tandem arrays where each island in the array has its own tDNA fragment and integrase gene. (Image (c) CM Hudson et. al, 2015)
TIGER uses the sequences 15kb to the left and right of the integrase gene midpoint to query against a BLAST database of species-specific reference genomes to search for query sequences in close proximity which signify the absence of an integrated genetic island and an aproximate attB sequence or, the site of putative island integration.
Graphical description of TIGER's usage of ping-pong BLAST for integrated genetic element (IGE) discovery. The corresponding regions of an IGE-bearing and uninterrupted reference genome pair (A) produce a sequence alignment pattern (B). Strand crossover presumably occurs with the direct repeat block (yellow). In TIGER (C), the first BLAST simultaneously locates the int-proximal end of the IGE and the attB, and the second locates the distal end of the IGE. (Image (c) CM Mageeney et. al, 2020)
TIGER2 improves on TIGER by enabling integrated genomic element identification in cases where the whole putative island sequence is not present on a singular scaffold/contig. Ultimately, the TIGER2 update introduces two new “split” modes that yield split GIs, in addition to the intact GIs. “CircleOrigin” mode finds split GIs that wrap around the origin of a circular replicon. “Cross” mode detects split GIs with termini on separate scaffolds in an incomplete assembly.
Graphical Demonstration of Fragmented Islands which Can be Identified with Tiger2. The same circular chromosome with 3 (colored) GIs is shown with a complete (A,B) or fragmented assembly (C). With complete assembly, if the origin of the linearized sequence of the circle is randomly chosen, it will occasionally fall within a GI, splitting the GI (B). Yields are shown for the various TIGER modes. The original mode can only find intact GIs on a single scaffold, while the new modes, CircleOrigin (applied to complete assemblies) and Cross (applied to fragmented assemblies), can additionally find the split islands. Because TIGER focuses on GI-flanking sequences, the Cross-mode call for a multiply split GI (red in panel C) will only include the terminal fragments and exclude middle GI fragments. (Image (c) CM Mageeney et. al, 2022)
TIGER requires the following programs to be available as system-wide executables and has been tested with the following software versions. We reccomend installing these packages and their dependencies using conda (instructions: Installation).
- Prokka v1.11 (https://github.com/tseemann/prokka)
- PfTools v3 (https://github.com/sib-swiss/pftools3)
- tRNAscan-SE 2.0 (https://github.com/UCSC-LoweLab/tRNAscan-SE)
- Perl IPC::Run3
This tutorial recommends that you have a working version of Anaconda or Miniconda (Miniconda is suggested: https://docs.anaconda.com/miniconda/) to properly install all dependencies.
#Create a conda environment for Tiger
conda create --name Tiger
#load your new environment
conda activate Tiger
# install prokka. Note, prokka has a known bug with conda that you must address by editing its code
# base. See instructions below.
conda install -c conda-forge -c bioconda -c defaults prokka
#setup/update prokka databases
prokka --setupdb
#Install pftools
conda install bioconda::pftools
#Install tRNA-Scan
conda install bioconda::trnascan-se
#Install Perl modules
conda install bioconda::perl-ipc-run3
#Install Git (if not already available on your system)
conda install anaconda::git
# download Tiger repository from GitHub
git clone https://github.com/sandialabs/TIGER.git
# Optional: We recomend adding the path of TIGER/bin to your bash profile to avoid having to call the
# full program during use. We always advise making a copy of your bash profile before making any
# edits as a typo may have severe consequences.
# ensure the execultables in the TIGER repository have executable permission on your system
cd <path to TIGER folder>/bin
chmod +x *
#install wget if not available on your system
conda install anaconda::wget
# get PFAM_A hmm database from NCBI: this is an OLD PFAM database (v35.0) if you want another
# version, you can change it in the ftp path below (note, this is the version that is suggested to
# use with TIGER per https://doi.org/10.3389/fbinf.2022.866850). This database must be installed to
# "Path to Tiger Folder"/db as PFAM-A.hmm
wget https://ftp.ebi.ac.uk/pub/databases/Pfam/releases/Pfam35.0/Pfam-A.hmm.gz -o <path to TIGER folder>/db/Pfam-A.hmm.gz
# install gzip if not already available on your system
conda install conda-forge::gzip
# uncompress PFAM database:
gunzip <path to TIGER folder>/db/Pfam-A.hmm.gz
#IMPORTANT: Fix a bug in Prokka:
# This bug has been reported in several forums without an easy solve given.
# What's happening is that the prokka main script is unable to correctly parse the version
# number of bioperl that it retrieves with conda (giving the error: ‘Argument "1.7.8" isn't
# numeric in numeric lt (<))’; it thus exits Prokka immediately. A "quick and ugly" fix is to
# hash out the lines in their code that are doing this check. So, in the prokka script
# (at "path to your conda distribution"/envs/"conda env name"/bin/prokka) add a hash
# (#) to the beginning of lines 256-259 like so:
"#my $minbpver = "1.006002"; # for Bio::SearchIO::hmmer3"
"#my $bpver = $Bio::Root::Version::VERSION;"
"#msg("You have BioPerl $bpver");"
"#err("Please install BioPerl $minbpver or higher") if $bpver < #$minbpver;"
# We recommend using either nano or vim to make these changes
The only requirement of Islander is an output directory containing a single genomic file in fasta format with a '.fa' subscript.
Islander Flags:
- '-outDir': Output directory. Default: same directory as GENOME_FASTA_FILE.
- '-tax': Taxonomic info for query genome. Enter file in outDir containing NCBI taxonomy string with a '.tax' subscript, or use B for Bacteria, A for Archaea, M for Mycoplasmatales/Entomoplasmatales, G for Gracilibacteria/candidate division SR1. Automatically sets -gencode. Default: B.
- '-gencode': Genetic code table (see NCBI). Default: 11
- '-nickname': Brief name for genome (as might be used to start a locus_tag).
- '-criterion': Basis for overlap resolution, 3 options: random, score (7-test false positive formula), deltaGC. Default = score.
- '-virus': Comma-separated list of entries assigned as viruses.
- '-complete': Consider genome complete and categorize replicons. Default: consider genome incomplete and call all entries contigs.
- '-force': Overwrite current output files. Default: leave existing files.
- '-cpu': Number of cpus to use. Default: 2.
- '-tateronly': Toggle to exit after running tater.pl annotator. Default: off.
- Additional options: '-help', '-version', '-verbose', '-authors', '-license'
Executing Islander Example:
From your Conda Environment:
conda activate Tigercd <path to out directory containing singular .fa file> ; islander.pl -verbose <fasta file name>Note: if you made the "path to TIGER"/bin a system-wide executable by enabling execute with chmod and adding the path to your bash profile, you will not need to specify perl (perl "path to TIGER"/bin/islander.pl) or use the path to the script in your launch command. If not, you will need to call the program as: cd "path to out directory containing singular .fa file" ; perl "path to TIGER"/bin/islander.pl -verbose
To rerun this program: delete genome.stats
TIGER requires a singular genomic or metagenomic file in fasta format ('.fa' subscript) in an output directory and a BLAST database of species specific reference genomes. See https://github.com/sandialabs/SmartDBs for instructions on generating this database. Alternatively, if you only need a BLAST database for a single/few species you can reach out to Kelly Williams (kpwilli at sandia dot gov) or Katie Mageeney (cmmagee at sandia dot gov) for assisance. Please provide the GTDB taxonomic name for your species as well as the link to either a DropBox or GoogleDrive Folder for the database to be deposited in.
TIGER Flags:
- '-fasta': Genomic fasta DNA sequence file.
- '-db': Blast database of reference genomes, absolute path.
- '-search': Search type. Specify island or IS. Default: island.
- '-tax': Taxonomic info for query genome. Enter name of a file containing NCBI taxonomy string, or use B for Bacteria, A for Archaea, M for Mycoplasmatales/Entomoplasmatales, G for Gracilibacteria/candidate division SR1. Automatically sets -gencode. Default: B.
- '-gencode': Genetic code table to use (see NCBI). Default: 11.
- '-nickname': Brief name for genome (as might be used to start a locus_tag).
- '-circle': Specify C if all genomic DNA sequences are circular, L if all DNAs are linear, or a filename for a tab-delimited file of query and circularity (eg. acc.vers[tab]circular/linear). Default: C.
- '-cross': Three options: intact, cross, or circleOrigin. Default: intact.
- '-complete': Consider genome complete and categorize replicons. Default: consider genome incomplete and call all entries contigs.
- '-qlen': Query length for islands. Default: 15000 (3000 is always used in the second pass test for IS's that rule out island artifacts).
- '-force': Overwrite current output files. Default: leave existing files.
- '-outDir': Output directory. Default: same directory as GENOME_FASTA_FILE.
- '-cpu': Number of cpus to use. Default: 1.
- Additional options: '-help', '-version', '-verbose', '-authors', '-license'
Executing TIGER Example:
From your Conda Environment:
conda activate Tigercd <path to out directory containing singular .fa file> ; tiger.pl -verbose -db <path to reference genome database and database prefix> -fasta <fasta file name>Note: if you made the "path to TIGER"/bin a system-wide executable by enabling execute with chmod and adding the path to your bash profile, you will not need to specify perl (perl "path to TIGER"/bin/tiger.pl) or use the path to the script in your launch command. If not, you will need to call the program as: cd "path to out directory containing singular .fa file" ; perl "path to TIGER"/bin/tiger.pl -verbose -db "path to reference genome database and database prefix" -fasta "fasta file name".
To rerun this program, delete 'genome.island.nonoverlap.gff'
TIGER and Islander are complementary software which each produce an island output. Two additional software are required to have the complete island output. Resolve merges TIGER and Islander output files into one. Typing takes the resolved file and adds additional information including what type (phage, integrative conjugative element, or unknown), and produced fasta and gff files for each GI.
From your Conda Environment:
conda activate Tigercd <path to out directory containing genome.island.merge.gff and islander.gff files>; resolve3.pl mixed lenient genome genome.island.merge.gff islander.gff cd <path to out directory containing resolve output>; typing.pl resolve3.gff There are numerous output files generated from TIGER and Islander pipeline. Many of these are only needed if you plan to rerun the software. This guide will describe the 4 main output files
TIGER
-
Genome.island.merge.gff: 9 column gff file. [1] DNA accession, [2] TIGER software, [3] island, [4] Left coordinate, [5] Right coordinate, [6] Support, [7] GI direction, [8] empty, [9] ID=GI name (nickname.length(kbp).site); brief= length(kbp).site; coord= GI location (note cross will have + with both contigs); compose=simple, cross, circle; lenth=size in bp; context=site of integration; flanks=integration site data; flip=; bitsum=; gnm=reference genome example ;crossover=length of crossover site; int= which integrase type (Y=tyrosine, S=serine) and location;mid=; side=; end0=; end1=; OL=; OR=; OU=; mobQ1=; mobQ2; IS=;ISoverlap=;transposon=;ISidentical=;q1=query of left size; q2=query for right size; q1identity=identity of query 1 on reference genome; q2identity= identity of query 2 on reference genome; isleLseq=attL sequence and flank; unintSeq=attB sequence from reference; isleRseq=attR sequence and flank; mean=false positive mean score;SD=false positive standard deviation; deltaint=delta integrase score; foreign=foreignness score; housekeep=housekeeping score; hypoth=hypothetical score; delta_GC=GC change score; dinuc=dinucleotide score; islrScore=false positive score for island; compScore=overall false positive score; taxid=gencode=genetic code; replicon=; qlen=query length; ints=integrase list; idok=check of ID name; gnmok=check genome; db=refernce database; reject= did this reject any GIs;positive=;status=relation to other GIs Islander
-
Islander.gff: 9 column gff file. [1] DNA accession, [2] island_finder, [3] genomic_island, [4] Left coordinate, [5] Right coordinate, [6] Score, [7] GI direction, [8] empty, [9] ID= length(kb).t(m)RNA site; trna= which tRNA is integrated; int_site=; int_site_type=; trna_dupe= L=; R=; tRNA_len= length of tRNA; tRNA_aaa=tRNA amino acid 3 letter code; tRNA_aa=tRNA 1 letter code; A_site=; J_site=; questionable=; qStart= query start; qEnd=query end; hitStart=GI start; hitEnd=GI end; percent_id=how similar query and hit; bit_score=; size=length of GI; group=; segment=; origin=; proximal=; distal=; isleLseq= attL sequence and flank; isleRseq= attR sequence and flank; deltaint=delta integrase score; foreign=foreignness score; housekeep=housekeeping score; hypoth=hypothetical score; delta_GC=GC change score; dinuc=dinucleotide score; intList=integrase list; intCoords= coordinates of integrase
Resolve
- resolve3.gff: 9 column gff file. [1] DNA accession, [2] Software that called GI, [3] island, [4] Left coordinate, [5] Right coordinate, [6] false positiveScore, [7] GI direction, [8] %support [9] ID=GI name (nickname.length(kbp).site); target= integration site; coord= GI location (note cross will have + with both contigs); compose=simple, cross, circle; isleLseq=attL sequence and flank; unintSeq=attB sequence from reference; isleRseq=attR sequence and flank; gnm=reference genome example; OL=; OU=; OR=; crossover=length of crossover site; idok=check of ID name; gnmok=check genome; db=reference database;OLL=;OLR=;ORL=;ORR=;center=; supp= support score; ints= which integrase type (Y=tyrosine, S=serine) and location;deltaside= which side the integrase is on ; Lenth=size in bp; deltaint=delta integrase score; foreign=foreignness score; housekeep=housekeeping score; hypoth=hypothetical score; delta_GC=GC change score; dinuc=dinucleotide score;overall=composite of false positive score; log score= calculated false positive score; rejectees=GIs rejected by this GI;rejected_by=was this rejected and by which GI;tandem=is this in a tandem GI; tandem_power=how many in tandem;tandem_pos=what number is this GI;
Typing
- islesFinal.gff: 9 column gff file. [1] DNA accession, [2] Software that called GI, [3] island, [4] Left coordinate, [5] Right coordinate, [6] false positiveScore, [7] GI direction, [8] %support [9] ID=GI name (nickname.length(kbp).site); supp= support score;intLongest=length of longest integrase amino acid; compose=simple, cross, circle; coord= GI location (note cross will have + with both contigs);crossover=length of crossover site;type=What kind of GI (phage,ICE,other); phage=phage score; ice=ICE score; source=which software; OLL=;OLR=;ORL=;ORR=; isleLseq= attL sequence and flank; unintSeq=attB sequence from reference; isleRseq=attR sequence and flank; gnm=reference genome example; ints= which integrase type (Y=tyrosine, S=serine) and location;deltaside= which side the integrase is on ; Lenth=size in bp; deltaint=delta integrase score; foreign=foreignness score; housekeep=housekeeping score; hypoth=hypothetical score; delta_GC=GC change score; dinuc=dinucleotide score; tandem=is this in a tandem GI; tandem_power=how many in tandem;tandem_pos=what number is this GI; idok=check of ID name; typeok=is this typed; gnmok=check genome; db=reference database;
Please leave an issue on this GitHub repo or reach out to Kelly Williams (kpwilli at sandia dot gov) or Katie Mageeney (cmmagee at sandia dot gov) for questions or assistance.
