Content summary:
NLRexpress is an open-source bundle of 17 machine learning-based predictors trained for detecting sequence motifs specific to plant NLR-associated domains and is organised in 4 distinct modules:
-
CCexpress: for an extended version of the EDVID motif: "RDbbbDbEDVID" (b - hydrophobic residue), which covers the 3rd alpha helice of the CC domain.
-
TIRexpress: comprises 6 predictors for the following conserved motifs found in plant TIR domains :
-
$\beta$ A: FLSFRGEDTR -
$\alpha$ A: KNFTSHL -
$\beta$ C: SRISIVVF -
$\alpha$ C: WCLDEL -
$\beta$ D$\alpha$ D1: VLPVFYDVDPSDVRKQ -
$\alpha$ D3: WREALTEVANLSG
-
-
NBSexpress: comprises 11 predictors for the following conserved motifs specific to plant NBS/NB-ARC domains :
- VG: bbGRx
- P-loop: GbGGbGKTT
- RNBS-A: FDbrhWhshs
- Walker-B: KRbbbbDD
- RNBS-B: KbbbTTR
- RNBS-C: LseeeSWeLF
- RNBS-D: CFLYCSLFP
- GLPL: GLPLA
- MHD: bHD
-
LRRexpress: for detecting "LxxLxL" motifs in Leucine-rich repeat (LRR) domains.
Not limited to plant NLRs, the LRRexpress module was developed to be also used for screening other LRR-containing protein classes regardless of their taxonomy.
The LRRexpress module is a much faster complementary alternative to LRRpredictor[1]. While LRRpredictor consists of a bundle of 8 voting estimators based on secondary structure predictions and sequence variability profiles generated on a large universal sequence database (Uniprot20), LRRexpress does not require the computation of secondary structure predictions and utilises a miniature targeted sequence database using JackHMMER [2][3] which significantly improves the speed of the pipeline with an average of less than 40s per NLR sequence (though average times might vary depending on computer specs).
For a limited number of protein sequences (under 100 seq), predictions can be run online on our webserver at https://nlrexpress.biochim.ro/
NLRexpress was tested only on Ubuntu 18.04, but older/newer distributions should be compatible.
git clone https://github.com/eliza-m/NLRexpress
cd NLRexpress
conda env create -f environment.yml
conda activate nlrexpress
If conda (anaconda/miniconda) is not installed, installation steps can be found at https://docs.conda.io/projects/conda/en/latest/user-guide/install/linux.html
The HMMER version 3.3 can be installed in Ubuntu 18 via:
sudo apt-get install hmmer
However, this might be superseded in the future by up-coming releases. The HMMER version archive can be found at : http://hmmer.org/download.html
wget https://nlrexpress.biochim.ro/datasets/models.tar.gz
tar -xf models.tar.gz
Now the installation is complete and NLRexpress ready for use.
The input file can contain multiple protein sequences in FASTA format. There is no constraint for the number of characters per line. The name of the query protein will be defined as the string beginning after the first '>' symbol until the first occurrence of a space character.
An example is provided with the repository. (Please note that the NLRexpress requires the conda environment to be active) :
While running a status log will be printed on the screen.
user@server:/home/user/git//NLRexpress$ conda activate nlrexpress
(nlrexpress) user@server:/home/user/git//NLRexpress$ python nlrexpress.py --input sample/input/zar1_rpp1.fa --outdir sample/output_ref/ --module all
22/08/2022 15:57:49: ############ NLRexpress started ############
22/08/2022 15:57:49: Input FASTA: sample/input/zar1_rpp1.fa
22/08/2022 15:57:49: Checking FASTA file - started
22/08/2022 15:57:49: Checking FASTA file - done
22/08/2022 15:57:49: Running JackHMMER - started
22/08/2022 15:58:48: Running JackHMMER - done
22/08/2022 15:58:48: Preparing features: Parsing HMM profile - started
22/08/2022 15:58:48: Running CCexpress : started
22/08/2022 15:58:49: Preparing features: NN input for motif extEDVID started
22/08/2022 15:58:49: Preparing features: NN input for motif extEDVID done
22/08/2022 15:58:49: Running CCexpress : Running extEDVID predictor: ...done
22/08/2022 15:58:49: Running CCexpress : done
22/08/2022 15:58:49: Running TIRexpress : started
22/08/2022 15:58:49: Preparing features: NN input for motif bA started
22/08/2022 15:58:49: Preparing features: NN input for motif bA done
22/08/2022 15:58:49: Running TIRexpress : Running bA predictor: ...done
22/08/2022 15:58:49: Preparing features: NN input for motif aA started
22/08/2022 15:58:49: Preparing features: NN input for motif aA done
22/08/2022 15:58:49: Running TIRexpress : Running aA predictor: ...done
22/08/2022 15:58:49: Preparing features: NN input for motif bC started
22/08/2022 15:58:50: Preparing features: NN input for motif bC done
22/08/2022 15:58:50: Running TIRexpress : Running bC predictor: ...done
22/08/2022 15:58:50: Preparing features: NN input for motif aC started
22/08/2022 15:58:50: Preparing features: NN input for motif aC done
22/08/2022 15:58:50: Running TIRexpress : Running aC predictor: ...done
22/08/2022 15:58:50: Preparing features: NN input for motif bDaD1 started
22/08/2022 15:58:50: Preparing features: NN input for motif bDaD1 done
22/08/2022 15:58:50: Running TIRexpress : Running bDaD1 predictor: ...done
22/08/2022 15:58:50: Preparing features: NN input for motif aD3 started
22/08/2022 15:58:50: Preparing features: NN input for motif aD3 done
22/08/2022 15:58:50: Running TIRexpress : Running aD3 predictor: ...done
22/08/2022 15:58:50: Running TIRexpress ...done
22/08/2022 15:58:50: Running NBSexpress : started
22/08/2022 15:58:50: Preparing features: NN input for motif VG started
22/08/2022 15:58:50: Preparing features: NN input for motif VG done
22/08/2022 15:58:51: Running NBSexpress : Running VG predictor: ...done
22/08/2022 15:58:51: Preparing features: NN input for motif P-loop started
22/08/2022 15:58:51: Preparing features: NN input for motif P-loop done
22/08/2022 15:58:52: Running NBSexpress : Running P-loop predictor: ...done
22/08/2022 15:58:52: Preparing features: NN input for motif RNSB-A started
22/08/2022 15:58:52: Preparing features: NN input for motif RNSB-A done
22/08/2022 15:58:52: Running NBSexpress : Running RNSB-A predictor: ...done
22/08/2022 15:58:52: Preparing features: NN input for motif RNSB-B started
22/08/2022 15:58:52: Preparing features: NN input for motif RNSB-B done
22/08/2022 15:58:53: Running NBSexpress : Running RNSB-B predictor: ...done
22/08/2022 15:58:53: Preparing features: NN input for motif RNSB-C started
22/08/2022 15:58:53: Preparing features: NN input for motif RNSB-C done
22/08/2022 15:58:54: Running NBSexpress : Running RNSB-C predictor: ...done
22/08/2022 15:58:54: Preparing features: NN input for motif RNSB-D started
22/08/2022 15:58:54: Preparing features: NN input for motif RNSB-D done
22/08/2022 15:58:54: Running NBSexpress : Running RNSB-D predictor: ...done
22/08/2022 15:58:54: Preparing features: NN input for motif Walker-B started
22/08/2022 15:58:54: Preparing features: NN input for motif Walker-B done
22/08/2022 15:58:55: Running NBSexpress : Running Walker-B predictor: ...done
22/08/2022 15:58:55: Preparing features: NN input for motif GLPL started
22/08/2022 15:58:55: Preparing features: NN input for motif GLPL done
22/08/2022 15:58:55: Running NBSexpress : Running GLPL predictor: ...done
22/08/2022 15:58:55: Preparing features: NN input for motif MHD started
22/08/2022 15:58:55: Preparing features: NN input for motif MHD done
22/08/2022 15:58:56: Running NBSexpress : Running MHD predictor: ...done
22/08/2022 15:58:56: Running NBSexpress ...done
22/08/2022 15:58:56: Running LRRexpress ...started
22/08/2022 15:58:56: Preparing features: NN input for motif LxxLxL started
22/08/2022 15:58:56: Preparing features: NN input for motif LxxLxL done
22/08/2022 15:58:56: Running LRRexpress : Running LxxLxL predictor: ...done
22/08/2022 15:58:56: Running LRRexpress ...done
22/08/2022 15:58:56: Printing final results...started
22/08/2022 15:58:56: Printing final results...done
22/08/2022 15:58:56: ############ NLRexpress finished ############
Customized parameters are described in help mode:
python nlrexpress.py --help
Usage: nlrexpress.py [OPTIONS]
Predict NLR-related motifs
Options:
--input TEXT Input FASTA file [required]
--outdir TEXT Output folder [required]
--module TEXT Predifined prediction modules :
- cc : CCexpress contains motif predictors for the CC domain:
extEDVID: rdhhhdhEDVID
- tir : TIRexpress contains motif predictors for the TIR domain:
bA: FLSFRGEDTR
aA: KNFTSHL
bC: SRISIVVF
aC: WCLDEL
bDaD1: VLPVFYDVDPSDVRKQ
aD3: WREALTEVANLSG
- nbs : NBSexpress contains motif predictors for the NBS/NBARC domain:
VG: bbGRx
P-loop: GbGGbGKTT
RNBS-A: FDbrhWhshs
Walker-B: KRbbbbDD
RNBS-B: KbbbTTR
RNBS-C: LseeeSWeLF
RNBS-D: CFLYCSLFP
GLPL: GLPLA
MHD: bHD
- lrr : LRRexpress contains motif predictors for the NBS/NBARC domain:
LRR pattern: LxxLxL
- all : All modules above: CC, TIR, NBS and LRR express [required]
--outformat TEXT Output format layout :
- short : Only the hits with more than 20% probability will be printed ( one line per motif )
- long : All predicted residues will be printed ( one line per residue )
- all [default] : Both short and long output formats are generated
--cpunum INTEGER Set the number of CPU threads to be used.
--writeinputfile BOOLEAN Write input file to be used in case of job intreruption.
--help Show this message and exit.
Prediction results will be found within the provided output directory path:
NLRexpress$ ls -l sample/output/
ls -l sample/output_ref/
-rw-rw-r-- 1 eliza eliza 965535 Aug 22 15:58 zar1_rpp1-1.hmm
-rw-rw-r-- 1 eliza eliza 965525 Aug 22 15:58 zar1_rpp1-2.hmm
-rw-rw-r-- 1 eliza eliza 2100 Aug 22 15:57 zar1_rpp1.fasta_proc
-rw-rw-r-- 1 eliza eliza 5360 Aug 22 15:58 zar1_rpp1.log
-rw-rw-r-- 1 eliza eliza 447768 Aug 22 15:58 zar1_rpp1.long.output.txt
-rw-rw-r-- 1 eliza eliza 7687 Aug 22 15:58 zar1_rpp1.short.output.txt
A sample output for zar1_rpp1.fa example can be found in sample/output_ref/ for comparison purposes.
Two types of output formats are provided: a summarising and a detailed version.
The summarizing (short) version filename.short.output.txt which contains from right to left: sequence name (from the FASTA header), the residue ID at which the predicted motif starts, the motif name, estimated probability, 5 positions upstream of the motif, the motif sequence and 5 positions downstream. Shown are only potential motifs yielding a probability estimate above 20%.
#ProtName ResId Motif Proba | -5pos | MotifSeq | +5pos |
Q38834_ZAR1 66 extEDVID 99.74 | LVADL | RELVYEAEDILV | DCQLA |
Q38834_ZAR1 160 VG 99.24 | YDHTQ | VVGLE | GDKRK |
Q38834_ZAR1 189 P-loop 100.00 | IMAFV | GMGGLGKTT | IAQEV |
Q38834_ZAR1 212 RNSB-A 99.81 | EIEHR | FERRIWVSVS | QTFTE |
Q38834_ZAR1 260 Walker-B 97.30 | QYLLG | KRYLIVMD | DVWDK |
Q38834_ZAR1 291 RNSB-B 98.46 | RGQGG | SVIVTTR | SESVA |
Q38834_ZAR1 318 RNSB-C 99.94 | HRPEL | LSPDNSWLLF | CNVAF |
Q38834_ZAR1 357 GLPL 99.98 | VTKCK | GLPLT | IKAVG |
Q38834_ZAR1 418 RNSB-D 98.75 | SHLKS | CILTLSLYP | EDCVI |
Q38834_ZAR1 487 MHD 99.65 | IITCK | IHD | MVRDL |
Q38834_ZAR1 512 LxxLxL 93.98 | PEGLN | CRHLGI | SGNFD |
Q38834_ZAR1 561 LxxLxL 99.73 | TDCKY | LRVLDI | SKSIF |
Q38834_ZAR1 588 LxxLxL 99.93 | ASLQH | LACLSL | SNTHP |
Q38834_ZAR1 612 LxxLxL 99.95 | EDLHN | LQILDA | SYCQN |
Q38834_ZAR1 636 LxxLxL 99.90 | VLFKK | LLVLDM | TNCGS |
Q38834_ZAR1 660 LxxLxL 54.88 | GSLVK | LEVLLG | FKPAR |
Q38834_ZAR1 686 LxxLxL 99.87 | KNLTN | LRKLGL | SLTRG |
...
The detailed long format filename.long.output.txt, contains one row per sequence residue and the corresponding probability estimate for each of the motif predictors. The first and last positions of the sequences do not have an estimate due to the fact that the predictors require 5 positions up- and down-stream the predicted position.
#ProtName resid aa extEDVID bA aA bC a bDaD1 aD3 VG P-loop RNSB-A RNSB-B RNSB-C RNSB-D Walker-B GLPL MHD LxxLxL
Q38834_ZAR1 1 M - - - -
- - - - - - - - - - - -
Q38834_ZAR1 2 V - - - -
- - - - - - - - - - - -
Q38834_ZAR1 3 D - - - -
- - - - - - - - - - - -
Q38834_ZAR1 4 A - - - -
- - - - - - - - - - - -
Q38834_ZAR1 5 V - - - -
- - - - - - - - - - - -
Q38834_ZAR1 6 V 0.00 0.00 0.00 0.00 0.0 0.00 0.00 0.04 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.09 0.00
Q38834_ZAR1 7 T 0.00 0.00 0.01 0.00 0.0 0.00 0.00 0.02 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00
Q38834_ZAR1 8 V 0.00 0.00 0.00 0.00 0.0 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Q38834_ZAR1 9 F 0.00 0.00 0.00 0.00 0.0 0.00 0.00 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Q38834_ZAR1 10 L 0.00 0.00 0.00 0.00 0.0 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.07 0.00
Q38834_ZAR1 11 E 0.00 0.00 0.00 0.00 0.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Q38834_ZAR1 12 K 0.00 0.00 0.00 0.00 0.0 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00
Q38834_ZAR1 13 T 0.00 0.00 0.00 0.00 0.0 0.00 0.00 0.02 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.02 0.10
Q38834_ZAR1 14 L 0.10 0.00 0.00 0.00 0.0 0.00 0.00 0.04 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.08
Q38834_ZAR1 15 N 0.00 0.00 0.00 0.00 0.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Q38834_ZAR1 16 I 0.00 0.00 0.00 0.00 0.0 0.00 0.00 0.18 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.01
Q38834_ZAR1 17 L 0.00 0.00 0.00 0.00 0.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Q38834_ZAR1 18 E 0.00 0.00 0.00 0.00 0.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Q38834_ZAR1 19 E 0.00 0.00 0.00 0.00 0.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
If you use LRRpredictor please cite:
[1] Martin EC, Sukarta OCA, Spiridon L, Grigore LG, Constantinescu V, Tacutu R, Goverse A, Petrescu AJ. LRRpredictor-A New LRR Motif Detection Method for Irregular Motifs of Plant NLR Proteins Using an Ensemble of Classifiers. Genes (Basel). 2020 Mar 8;11(3):286. doi: 10.3390/genes11030286.
[2] Jaina Mistry, Robert D. Finn, Sean R. Eddy, Alex Bateman, Marco Punta, Challenges in homology search: HMMER3 and convergent evolution of coiled-coil regions, Nucleic Acids Research, Volume 41, Issue 12, 1 July 2013, Page e121, https://doi.org/10.1093/nar/gkt263.
[3] Robert D. Finn, Jody Clements, Sean R. Eddy, HMMER web server: interactive sequence similarity searching, Nucleic Acids Research, Volume 39, Issue suppl_2, 1 July 2011, Pages W29–W37, https://doi.org/10.1093/nar/gkr367