LocNES is a Support Vector Machine (SVM) predictor that locates classical nuclear export signals (NESs) in CRM1 cargoes. Export-Karyopherinβ CRM1 recognizes hundreds of broadly functioning proteins. Most CRM1 cargoes contain the classical NES (also known as leucine-rich NESs), a peptide with 8-15 amino acids, regularly spaced with conserved hydrophobic residues.

User input

Users submit either a single protein sequence or multiple protein sequences in FASTA format. The query sequence can also be uploaded in a file.

How LocNES works

LocNES first scans the query protein to gather all peptides that fit a modified version of the Kosugi NES consensus patterns. These peptides are called NES candidates. Each NES candiate has 15 amino acids (or less if located at the N-terminus). Next, the position specific scoring matrix (PSSM) score is computed for each candidate and all candidates are ranked according to the PSSM score. In addition, DISOPRED is used to calculate the disorder propensity of the query protein. The PSSM score rank, amino acid sequence, disorder propensity, and NES consensus patterns of the NES candidate are used to construct the SVM feature set. LocNES is implemented by LIBSVM 3.18.

LocNES output

The output of LocNES consists of four columns. The first column is the protein name; the second column is the location of the NES candidate; the third column is the sequence of the NES candiate; and the last column is the probability of the candiate being a real NES. Cross-validation showed that a cutoff of 0.1 gives 68% recall rate with 26% precision. Recall is defined as the fraction of real NESs whose probability score is higher than the cutoff value. Precision measures the percentage of real NESs among NES candidates with probability score higher than a cutoff.


If you use LocNES, please cite:

"LocNES: A computational tool for locating classical NESs in CRM1 cargo proteins."
Xu D., Marquis K., Pei J., Grishin N.V. and Chook, Y.M. (2014) Bioinformatics submitted

Contact Us

If you have questions, report bugs, or need stand-alone linux version, please email Darui Xu at Dr. YuhMin Chook's lab