Dragon PolyA Spotter (DPS)
 
Ver:1.200
 [since 2013-Feb-06 16:21:42 +0300]
 
Home | View jobs | Usage instructions | Team | Code

Algorithm, Code and Software

The current "Dragon PolyA Spotter" contains two types of predictive models. One type is based on Artificial Neural Networks (ANNs) and the other is based on Random Forests (RFs). ANN models are implemented in C++ programming language. RF models are implemented as WEKA models.


The ANN models can be installed on user machines, while RF models cannot. To install and run the ANN code do the following:

  • Build the executables from the source code (download this file).

  • Decompress the Poly(A)signalTesting-C.tar.bz2. This will create a directory named "Poly(A)signalTesting-C".

  • Go to the newly created "Poly(A)signalTesting-C" directory and run the command below.
    g++ -o polyASignalTesting src/PolyAsignalTesting-C.cpp;

  • The file "polyASignalTesting" is now your executable file. To execute the file "polyASignalTesting" run the following command:

    ./polyASignalTesting Data/ <path to fasta sequence file> logFile

    where <path to fasta sequence file> is the full path to the file containing the fasta sequences you want to analyze.


Data used in building and testing the model can be found here.


© 2011 King Abdullah University of Science and Technology. All rights reserved