Leucine-rich repeat (LRR) database and LRRfinder tool from the Royal Veterinary College (RVC)

LRRfinder is a web-based tool for the identification of leucine-rich repeats (LRRs) in protein sequences. Using our Toll-like receptor LRR database (tLRRdb), LRRfinder provides a simple web-accessible application for rapid, accurate LRR detection with a user-friendly output.

LRRs form key motifs in pathogen-associated molecular pattern (PAMP) binding sites and are present in a number of proteins co-ordinating vital cell processes. Their detection and analysis is a considerable step towards understanding protein-protein interactions within this protein family. LRRfinder is derived from a large database of unique, naturally occurring LRRs (tLRRdb) allowing the identification of not only highly conserved LRR sequences but also those which are more distinct from the commonly described LxxLxLxxN/CxL consensus.

LRRfinder is derived from the highly conserved (LxxLxLxxN/CxL) segments of LRRs which have been stored in the tLRRdb; 2651 unique, naturally occurring sequences with potential for expansion. Using both this novel LRR specific detection method and the tLRRdb, it is possible to find previously identified and unique LRRs which not only conform to the highly conserved LRR HS but also irregular LRRs stored in the database. It is the identification of these irregular LRRs and their specificity which makes LRRfinder to preferable previous detection methods for many key proteins.

A Position-Specific Scoring Matrix (PSSM) was created as a repre-sentative of the LRR-HS amino acid distributions to obtain their most statistically probable locations within a given sequence. The PSSM was constructed to form a probability matrix, derived from LRR-HS sequences stored in the LRRdb. A null PSSM is created from the user's sequence and represents the naturally occurring amino acid frequencies in the submitted sequence.

It is from these PSSMs that scores and significance values are derived. Results are given in a simple tabulated format with significance boundaries set by the user influencing the differentiation between significant and insignificant LRR predictions. For TLRs it is possible to search for other region matches such as the LRRNT, LRRCT and TIR domains.

This work was funded by a grant from the BBSRC (BB/D524040/1) and supplementary data can be found here.