1. About mirnaDetect:

A java based mining tool mirnaDetect is developed for detect potential pre-miRNAs from the genome-scale data. This program is based both on search and machine learning algorithms.  Because of the use of a machine learning algorithm in the program, one can adjust the threshold to get predicted sequences with different confidence. As is expected, higher threshold yields fewer sequences as pre-miRNAs, but has higher confidence. 

 

2. Preparation:

1. Make sure you have installed java in your computer. If not, please download: here

2. Download the package to anywhere in your computer. The source code is here: here

3. Uncompress the package: mirnaDetect_SourceCode.zip

4. Before running the program, please Read readme.txt.

 

3. Online Server—miRNApre:

This online pre-miRNA classification system is written based on our proposed machine learning algorithm. Using this online server, one can predict the given sequences as real pre-miRNAs or not. The online service of our classifier method is available (here).

 

4. Feature Extraction Algorithm

In the software ¡°mirnaDetect¡±, we propose a novel feature extraction algorithm that represents each RNA sequence with a 98-D feature vector. The proposed feature extraction algorithm is written with JAVA language and has been packaged in a JAR.  Users can download the JAR package (here), and Run the JAR package with the command ¡°Java -jar Extract98FeaturesForMiRNA.jar  YourOwnFile¡± , where YourOwnFile means the fasta file you want to extract features from.

 

5. Datasets:

The datasets used in our paper can be downloaded below:

(1) training set and testing set (here).

(2) the high-quality negative set(here).

(3) the feature set (here).

 

6. Citation:

1. Wei L, Liao M, Gao Y, Ji R, He Z, Zou Q. Improved and promising identification of human microRNAs by incorporating a high-quality negative set[J]. IEEE/ACM Transactions on Computational Biology and Bioinformatics, 2014, 11(1): 192-201. 

2. Wei L, Huang Y, Qu Y, Jiang Y, Quan Z. Computational analysis of miRNA target identification[J]. Current Bioinformatics, 2012, 7(4): 512-525.

 

7. Contact us£º

If you have any problem or advice about mirnaDetect, please contact us. Email adress is presented below,

Me: weileyi@stu.xmu.edu.cn OR  Prof. Zou : zouquan@nclab.net

 

ÖÐÎÄ°æʹÓÃ˵Ã÷

 

Leyi Wei

2016. 1