Identification of NBS-LRR genes in genomic sequences v1.2.4
This was created as part of the project HealthyGrape2 at the INRA.
[email protected] (original)
[email protected] (current)
NLGenomeSweeper is a command line pipeline that searches a genome for NBS-LRR (NLR) disease resistance genes based on the presence of the NB-ARC domain using the consensus sequence of the Pfam HMM profile (PF00931) and class specific consensus sequences built from Vitis vinifera. This pipeline can be used with a custom NB-ARC HMM consensus protein sequence(s) built for a species of interest or related species for greater power, separately for each type of NBS-LRR (TNLs, CNLs, NLs) and combine them into a single fasta file for use. This pipeline shows high specificity for complete genes and structurally complete pseudogenes. However, candidate regions are identified but may not necessarily represent functional genes and does not itself do gene prediction. A domain identification step is also included and the output in gff3 format can be used for manual annotation of NLR genes. Therefore, it is primarily for the identification of NLR genes for a genome where either no annotation exists or a large number of genes are expected to be absent due to repeat masking and difficulties in annotation. For many genomes this may be the case.
If you use NLGenomeSweeper for published research, please cite the paper.
Toda, N.; Rustenholz, C.; Baud, A.; Paslier, M.-C.L.; Amselem, J.; Merdinoglu, D.; Faivre-Rampant, P. NLGenomeSweeper: A Tool for Genome-Wide NBS-LRR Resistance Gene Identification. Genes 2020, 11, 333.
Thanks!
NLGenomeSweeper can be obtained from the GitHub page
This is a pipeline to be run on unix based machines. The following software must be available in your path. At least 2G of free memory is needed for InterProScan but more memory is better.
- Python 3, version 3.5 or greater
- blast+
- MUSCLE aligner (some newer versions are incompatible, 3.8.1551 is known to be compatible)
- SAMtools
- bedtools
- HMMER
- InterProScan with PANTHER database
- TransDecoder
Example commands for setting up an environment to run NLGenomeSweeper using conda and common utilities. This requires approximately 90 gibibytes of free disk space.
# create a new environment and install needed pacakages from conda
conda create -y -n NLGenomeSweeper -c bioconda -c conda-forge \
python=3.6 blast muscle=3.8.1551 samtools bedtools hmmer transdecoder openjdk
conda activate NLGenomeSweeper
# install interproscan with Panther
wget ftp://ftp.ebi.ac.uk/pub/software/unix/iprscan/5/5.45-80.0/interproscan-5.45-80.0-64-bit.tar.gz
tar -xzf interproscan-5.45-80.0-64-bit.tar.gz
ln -s $(pwd)/interproscan-5.45-80.0/interproscan.sh $(dirname $(which python))/interproscan
wget ftp://ftp.ebi.ac.uk/pub/software/unix/iprscan/5/data/panther-data-14.1.tar.gz
tar -xzf panther-data-14.1.tar.gz -C interproscan-5.45-80.0/data/
# download NLGenomeSweeper and install
git clone https://github.com/ntoda03/NLGenomeSweeper.git
ln -s $(pwd)/NLGenomeSweeper/NLGenomeSweeper $(dirname $(which python))/NLGenomeSweeper
NLGenomeSweeper -h
usage: NLGenomeSweeper [-h] -genome [-consensus ]
[-overwrite [T/F]] [-outdir ] [-t ]
optional arguments:
-h, --help show this help message and exit
Required arguments:
-genome A genome sequence in fasta format to search.
Input arguments (optional):
-consensus Also search the genome using a custom NB-ARC consensus sequence(s).
This is a protein fasta file of the domain sequence(s) only.
NOT RECOMMENDED. Species specific consensus sequences are built by default.
Output arguments (optional):
-overwrite [T/F] Whether to overwrite output files if they already exist. [Default F]
-outdir Path where output directory NLGenomeSweeper will be created. [Default ./]
Program control:
-t The number of threads to use.
Search for NB-ARC domains in the genome:
NLGenomeSweeper -genome Mrot_genome.fa
Search using custom consensus sequences:
NLGenomeSweeper -genome Mrot_genome.fa -consensus VvinTNL.fa
Here is an overview of the important output files created by the script. The files will be output by default in a folder called NLGenomeSweeper which will be created in the current working directory or else the user can specify and output location.
The primary output files of NLGenomeSweeper will be written to the output directory. Warning! A potential structural classification is given but these must be verified by manual annotation.
All_candidates.bed
This is the list of positions in the genome that were identified as probable NBS-LRR genes in bed format. Coordinates refer to the position of the NB-ARC domain and not of the potential gene. A potential classification is given.
Final_candidates.bed
This is the list of positions in the genome that were identified as probable NBS-LRR genes in bed format that has passed the filter for the presence of LRRs. Coordinates refer to the position of the NB-ARC domain and not of the potential gene. A potential classification is given.
Filtered_candidates.bed
This is the list of positions in the genome that were identified as probable NBS-LRR genes in bed format but filtered out because they did not contain an LRR. Coordinates refer to the position of the NB-ARC domain and not of the potential gene. A potential classification is given.
All_candidates.gff3
This is the annotation gff3 file based on the results of InterProScan to predicts domains and ORFs in the at the positions of predicted NBS-LRR genes.
The main pipeline has been tested on diverse plant species with agricultural or ecological importance, but the performance may suffer if you are looking at a species that is less well studied and with a long divergence time relative to what has been tested. CustomProfiler.py is a helper script to create a custom HMM profile and consensus sequence based on high quality known NBS-LRR genes from any species that can then be used by the main pipeline. It will isolate the NB-ARC domain from a file containing protein sequences of NRL genes. It is strongly suggested that if multiple classes are available (TNL, CNL, RNL) that each class is built separately and the consensus sequences aggregated into a single fasta file for use. By default this uses the Vitis vinifera CNL, NL, TNL, and RNL consensus domains. If you are working with a species very distant from this then it is suggested that you use the consensus sequences from a more closely related species as the reference domain or if that is not available then extract the NB-ARC domain manually from a known NLR gene in a more closely related species.
usage: CustomProfiler.py [-h] -verified -prefix
[-overwrite [T/F]] [-outdir ]
optional arguments:
-h, --help show this help message and exit
Required arguments:
-verified A fasta file containing protein sequences of known NBS-LRR genes.
-prefix Prefix for output files.
Domain arguments:
-reference_nbarc
A reference NB-ARC domain protein sequence for comparison in fasta format. [Default data/Vitis_vinifera_NB-ARC_consensus.fa]
Output arguments:
-overwrite [T/F] Whether to overwrite output files if they already exist. [Default F]
-outdir Set an output directory for program output. [Default ./NLGenomeSweeper]
Generate a custom HMM profile and consensus sequence:
python CustomProfiler.py -verified vvinifera_TNLgenes.fa -prefix VvinTNL
The primary output files of CustomProfiler will be written to the folder 00_profile_creation within the output directory.
00_profile_creation/<prefix>.fa
The consensus NB-ARC domain sequence generated for the known NBS-LRR genes.
00_profile_creation/<prefix>.hmm
The NB-ARC domain HMM profile generated for the known NBS-LRR genes.