Background
Many effector proteins from oomycete plant pathogens for manipulating the host have been found to contain a signal peptide followed by a conserved RXLR motif (Arg, any amino acid, Leu, Arg), and then sometimes EER (Glu, Glu, Arg). There are striking parallels with the malarial host-targeting signal (Plasmodium export element, or "Pexel" for short).
What it does
Takes a protein sequence FASTA file as input, and produces a simple tabular file as output with one line per protein, and two columns giving the sequence ID and the predicted class. This is typically just whether or not it had the selected RXLR motif (Y or N).
Bhattacharjee et al. (2006) RXLR Model
Looks for the oomycete motif RXLR as described in Bhattacharjee et al. (2006).
Matches must have a SignalP Hidden Markov Model (HMM) score of at least 0.9, a SignalP Neural Network (NN) predicted cleavage site giving a signal peptide length between 10 and 40 amino acids inclusive, and the RXLR pattern must be after but within 100 amino acids of the cleavage site. SignalP is run truncating the sequences to the first 70 amino acids, which was the default on the SignalP webservice used in Bhattacharjee et al. (2006).
Win et al. (2007) RXLR Model
Looks for the protein motif RXLR as described in Win et al. (2007).
Matches must have a SignalP Hidden Markov Model (HMM) score of at least 0.9, a SignalP Neural Network (NN) predicted cleavage site giving a signal peptide length between 10 and 40 amino acids inclusive, and the RXLR pattern must be after the cleavage site and start between amino acids 30 and 60. SignalP is run truncating the sequences to the first 70 amino acids, to match the methodology of Torto et al. (2003) followed in Win et al. (2007).
Whisson et al. (2007) RXLR-EER with HMM
Looks for the protein motif RXLR-EER using the heuristic regular expression methodolgy, which was an extension of the Bhattacharjee et al. (2006) model, and a HMM as described in Whisson et al. (2007).
All the requirements described above for Bhattacharjee et al. (2006) apply, but rather than just looking for RXLR with the regular expression R.LR the more complicated regular expression R.LR.{,40}[ED][ED][KR] is used. This means RXLR (Arg, any amino acid, Leu, Arg), then a stretch of up to forty amino acids before Glu/Asp, Glu/Asp, Lys/Arg. The EER part of the name is perhaps misleading as it also allows for DDR, EEK, and so on.
Unlike Bhattacharjee et al. (2006) which used the SignalP webservice which defaults to truncating the sequences at 70 amino acids, Whisson et al. (2007) used the SignalP 3.0 command line tool with its default of not truncating the sequences. This does alter some of the scores, and also takes a little longer.
Additionally HMMER 2.3.2 is run to look for a cross validated HMM for the RXLR-ERR domain based on known positive examples. There are no restrictions on where within the protein the HMM match must be found.
Note
Both Bhattacharjee et al. (2006) and Win et al. (2007) used SignalP v2.0, which is no longer available. The current release is SignalP v3.0 (Mar 5, 2007), so this is used instead. SignalP is called with the Eukaryote model and the short output (one line per protein). Any sequence truncation (e.g. to 70 amino acids) is handled via the intemediate sequence files.
References
If you use this Galaxy tool in work leading to a scientific publication please cite Cock et al. (2013) and the appropriate method paper(s):
Peter J.A. Cock, Björn A. Grüning, Konrad Paszkiewicz and Leighton Pritchard (2013). Galaxy tools and workflows for sequence analysis with applications in molecular plant pathology. PeerJ 1:e167 https://doi.org/10.7717/peerj.167
Stephen C. Whisson, Petra C. Boevink, Lucy Moleleki, Anna O. Avrova, Juan G. Morales, Eleanor M. Gilroy, Miles R. Armstrong, Severine Grouffaud, Pieter van West, Sean Chapman, Ingo Hein, Ian K. Toth, Leighton Pritchard and Paul R. J. Birch (2007). A translocation signal for delivery of oomycete effector proteins into host plant cells. Nature 450:115-118. https://doi.org/10.1038/nature06203
Joe Win, William Morgan, Jorunn Bos, Ksenia V. Krasileva, Liliana M. Cano, Angela Chaparro-Garcia, Randa Ammar, Brian J. Staskawicz and Sophien Kamoun (2007). Adaptive evolution has targeted the C-terminal domain of the RXLR effectors of plant pathogenic oomycetes. The Plant Cell 19:2349-2369. https://doi.org/10.1105/tpc.107.051037
Souvik Bhattacharjee, N. Luisa Hiller, Konstantinos Liolios, Joe Win, Thirumala-Devi Kanneganti, Carolyn Young, Sophien Kamoun and Kasturi Haldar (2006). The malarial host-targeting signal is conserved in the Irish potato famine pathogen. PLoS Pathogens, 2(5):e50. https://doi.org/10.1371/journal.ppat.0020050
Trudy A. Torto, Shuang Li, Allison Styer, Edgar Huitema, Antonino Testa, Neil A.R. Gow, Pieter van West and Sophien Kamoun (2003). EST mining and functional expression assays identify extracellular effector proteins from the plant pathogen phytophthora. Genome Research, 13:1675-1685. https://doi.org/10.1101/gr.910003
Sean R. Eddy (1998). Profile hidden Markov models. Bioinformatics, 14(9):755–763. https://doi.org/10.1093/bioinformatics/14.9.755
Nielsen, Engelbrecht, Brunak and von Heijne (1997). Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Protein Engineering, 10:1-6. https://doi.org/10.1093/protein/10.1.1
This wrapper is available to install into other Galaxy Instances via the Galaxy Tool Shed at http://toolshed.g2.bx.psu.edu/view/peterjc/tmhmm_and_signalp