LRRpredictor

General Info

LRRpredictor is an open-source tool for detecting LRR motifs within leucine rich repeats proteins. It resides on secondary structure, relative solvent accessibility and disorder predictions that are performed using RaptorX-Property [1-4] and sequence variability profiles generated using HH-suite [5,6] on Uniprot20 sequence database.

Run locally

LRRpredictor source can be found on GitHub at:
https://github.com/eliza-m/LRRpredictor_v1

Also provided are a Dockerfile from which a docker image can be built and a pre-installed Docker image can be pulled from our docker repository.

Detailed installation instruction are provided in README.md file.

Usage

At the moment, only one single sequence can be processed in a single job.

Output files

1. Short output

"ProteinName.predshort.txt"

Displays only the potential LRR motifs identified by LRRpredictor, that yielded a probability value over 0.5 (if any).
Probabilities are also shown for each classifier (clf 1-8).
If none LRR motif was detected this file will contain only the header.

Example:

#Protein

pos

clf1

clf2

clf3

clf4

clf5

clf6

clf7

clf8

LRRpred

-5

-4

-3

-2

-1

+10

gpa216045

864

0.912

0.949

1.000

0.778

0.898

0.893

0.999

0.682

0.889

Header description:
* prot           - protein name
* pos            - residue number where a detected LRR motif starts (i.e first `L` from `LxxLxL` minimalistic motif)
* clf1-8         - Each classifiers predicted probability (min: 0, max 1)
* LRRpred    - LRRpredictor probability based on all eight classifiers.

Starting from columns 12 until the end, the amino acid sequence of the detected LRR motif is shown: 5 positions upstream the motif (-5 to -1), the minimalistic motif 'LxxLxL' and 5 positions downstream (6 to 10).

2. Long output

"ProteinName.pred.txt"

Example:

#Protein	resid	aa	clf1	clf2	clf3	clf4	clf5	clf6	clf7	clf8	LRRpred
gpa216045	864	L	0.912	0.949	1.000	0.778	0.898	0.893	0.999	0.682	0.889

Header description:
* prot               - protein name
* resid              - residue number
* aa                  - amino acid one letter code
* unused          - unused field (this field is used only training and testing data and indicates the position where a true LRR motifs starts; these positions were identified from structural files).
* clf1-8             - Each classifiers predicted probability (min: 0, max 1)
* LRRpredictor - LRRpredictor probability based on all eight classifiers.

3. Data used as input

"ProteinName.input"

Data used as input for LRRpredictor - RaptorX-property SS, RSA, disorder predictions and variability profile.

Reference

If you use LRRpredictor please cite:

Eliza C. Martin, Octavina C. A. Sukarta, Laurentiu Spiridon, Laurentiu G. Grigore, Vlad Constantinescu, Robi Tacutu, Aska Goverse, Andrei-Jose Petrescu. LRRpredictor - a new LRR motif detection method for irregular motifs of plant NLR proteins using ensemble of classifiers. Genes 2020, 11, 286.

Click to view paper

Bibliography

[1] Wang, S.; Li, W.; Liu, S.; Xu, J. RaptorX-Property: a web server for protein structure property prediction. Nucleic Acids Res. 2016, 44, W430-W435.

[2] Wang, S.; Peng, J.; Ma, J.; Xu, J. Protein Secondary Structure Prediction Using Deep Convolutional Neural Fields. Sci. Rep. 2016, 6, -11.

[3] Wang, S.; Ma, J.; Xu, J. AUCpreD: Proteome-level protein disorder prediction by AUC-maximized deep convolutional neural fields. In Proceedings of the Bioinformatics; Oxford University Press, 2016; Vol. 32, pp. i672-i679.

[4] Wang, S.; Sun, S.; Xu, J. AUC-maximized deep convolutional neural fields for protein sequence labeling. In Proceedings of the Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Springer Verlag, 2016; Vol. 9852 LNAI, pp. 1-16.

[5] Remmert, M.; Biegert, A.; Hauser, A.; Soding, J. HHblits: Lightning-fast iterative protein sequence searching by HMM-HMM alignment. Nat. Methods 2012, 9, 173-175.

[6] Steinegger, M., Meier, M., Mirdita, M., Vohringer, H., Haunsberger, S. J., Soding, J. HH-suite3 for fast remote homology detection and deep protein annotation. BMC Bioinformatics 2019, 473. doi: 10.1186/s12859-019-3019-7

Contact

For any issue or suggestion please feel free to write us at :

e-mail: eliza.martin@biochim.ro