Download Latest Version negative_examples.zip (30.0 MB)
Email in envelope

Get an email when there's a new version of sourcesinc

Home / cplxmirna
Name Modified Size InfoDownloads / Week
Parent folder
cplxmirna_data_MiPred.tar.xz 2019-09-28 3.3 MB
0
README.md 2019-03-22 1.8 kB
0
cplxmirna-0.5.tar.gz 2019-03-20 28.3 MB
0
Totals: 3 Items   31.7 MB 0

This is a distribution of the source code used in:

Jonathan Raad, Georgina Stegmayer and Diego H. Milone, "Complexity measures of the mature for improving pre-miRNAs prediction"

The last version of the source code can be found at: http://sourceforge.net/projects/sourcesinc/files/cplxmirna/

Contact:

This source code generates the results obtained in the article. There is also a dataset that includes Permutation Entropy (PE) and Lempel-Ziv (LZ) features (pre-calculated). The feature based on Levenshtein distance (LD) is calculated in real time.

To use this code, please run the ‘main.m’ script. The script was tested with Matlab R2014 and Ubuntu 16.04. The main script contains 2 inputs parameters. The first parameter is the name of the feature ('SF', 'LD', 'PE', 'LZ', 'ALL'). The second parameter is the classifier ('KNN' or 'DNN').

For example, run:

$ main('LZ', 'KNN');

This function creates log files in the "log" folder and automatically names them with time-stamps.

Installation

If you want to use the LD feature, the following software must be installed:

Directory structure

In the root directory you can find these folders:

  • data: contains the dataset
  • log: used to save the output files.
  • models: contains the machine learning models.
  • utils: contains all the functions of the code.
  • src: contains the function to calculate the LD, PE and LZ.

Additionally:

  • main.m: the main function of the source code.
  • traintest.m: the cross-validation for training and test scripts.
  • README: the file that you are reading.
Source: README.md, updated 2019-03-22