Prediction of lncRNA-protein interactions


This is a website for predicting the interaction between long noncoding RNAs and proteins. By coding RNA and protein sequences into vectors, we use matrix multiplication to give score to each RNA-protein pair. This score can be the measurement of interactions between the RNA-protein pair. Comparing to existing approaches, this method shortens the time for training matrix. It also theoretically guarantees the results to be the best solution. The method has shown good ability to discriminate interacting/non-interacting RNA-protein pairs and to predict the RNA-protein interaction within a given complex.


  1. We code the protein and RNA sequences into vectors by using information of secondary structure, hydrogen-bonding propensities, and Van der Waal’s interaction. Fourier’s Series is used to guarantee all the vectors have the same dimension.
  2. Fisher Linear Discrimnant Method is used to compute the main matrix. The theory of Fisher Method guarantees the main matrix is theoretically the best basing on the model we use.
  3. We use matrix computation (accompanied with some adjustments of score range) to get the score for each aspect of information.
  4. Take the arithmetic mean of scores from different aspects of information to be the final measurement of interaction.


Firstly, 4-fold cross validation was used to validate the ability of discriminating interactive and non-interactive pairs.

Secondly, we tested the method in Human MRP, RNase P, and HOTAIR complex. Our method can predict the interactions within given complex.

Thirdly, our method predicted nuclear proteins tend to interact with lncRNAs, which is in agreement of known fact.

Detail data and results can be found in paper

Welcome to lncPro

- (input a single RNA and multiple protein sequences)


Enter RNA Sequence in FASTA format: example


Upload Protein Sequences file in FASTA format: example file (download)

Download the package for
Download the package for lncPro (Mac).download.png

Citation: Lu et al.: Computational prediction of associations between long non-coding RNAs and proteins. BMC Genomics 2013 14:651. doi:10.1186/1471-2164-14-651