non-canonical MicroRNA Target (ncMirTar) predictor

About

The server is designed for prediction of non-canonical microRNA targets, which have fewer than seven Watson-Crick base pairs with nucleotide 1-8 from 5' end of the microRNA. ncMirTar only requires the knowledge of the microRNA and mRNA sequences.

This page provides a pre-computed database of miRNA-mRNA interactions and a web server that offers the prediction of targets for novel microRNAs. The database includes predicted non-canonical targets for known microRNA in human and mouse and allows retrieving predictions by microRNA id, mRNA gene or transcript id, or both. The web server predicts non-canonical targets in the human or mouse genome of the given microRNA sequence which is a potential new microRNA and is not deposited in the miRBase.

For a given query that may concern multiple miRNAs and mRNAs, the results of retrieval from the database and from the ncMirTar web server include a list of canonical mRNA targets for prediction with other methods (see the list below) and detailed predictions for the non-canonical targets. These predictions include putative interaction sites and propensities of these putative interactions at the site and targets gene level. We note that each considered target is categorized either as canonical (if it has at least one canonical site) or non-canonical (if all its sites are non-canonical).

Canonical miRNA target predictions can be retrieved through other web servers:

ncMirTar Release v1.0
Genome	# of microRNA¹	# of mRNA genes²	# of mRNA transcripts²
Human	2588	19135	37326
Mouse	1915	19964	27963

¹ microRNAs are from miRBase Release June 2014
² mRNA genes and transcripts are from RefSeq Release November 2014

The executive files of ncMirTar are available at http://github.com/BiomineLab/ncMirTar/.

Search targets from the database

Please follow the three steps below to search targets from the database:

1. Enter either microRNA identifier or mRNA identifier, or both of them

Search by microRNA identifier:

Searching accepts the microRNA identifier from miRBase. Only -3p or -5p arm is acceptable. The * format is not allowed.

Search by mRNA identifier:

Searching accepts the mRNA accession id from RefSeq, or the mRNA symbol from Genbank. The gene name is case sensitive.

2. Provide your e-mail address (required):

Please provide your e-mail address to be notified when results are ready.

3. Search:

Predict targets using ncMirTar

Please follow the four steps below to make prediction:

1. Select species

2. Enter the microRNA sequence

Enter the microRNA sequence in a line in the following text field. Predictor accepts only one microRNA sequence at each time. Predictor accepts microRNA sequence from 5' end to 3' end, and the microRNA size should range between 15 and 30.

3. Provide your e-mail address (required):

Please provide your e-mail address to be notified when results are ready.

4. Predict:

Please note the prediction for a new microRNA sequence will take up to three hours.

Materials

microRNA list from Release 21 of miRBase for Human and Mouse
mRNA gene list from Release Nov 2014 of Genbank for Human and Mouse

TRAINING - Training dataset used in the paper
TEST_gene - Gene-level dataset used in the paper
TEST_expression - Gene-level at transcriptome/proteome scale dataset used in the paper
TEST_duplex - Duplex-level dataset used in the paper

References

Upon the usage the users are requested to use the following citation:

Fan X., Wang C., Kurgan L. Accurate high-throughput prediction of non-canonical microRNA targets. Submitted

Acknowledgments

We acknowledge with thanks the following software used as a part of this server:

BLAST - Finding regions of similarity between biological sequences
ViennaRNA - Programs for the prediction and comparison of RNA secondary structures.