Difference between revisions of "Profppikernel"

From Rost Lab Open
Line 7: Line 7:
   
 
== Availability ==
 
== Availability ==
We have included both executables in a Debian package, together with man pages and a sample classification problem. There are two common ways to install it.
+
We have included both executables in a .tar.gz source package, together with a sample classification problem. Download it [http://www.rostlab.org/~hampt/profppikernel-1.0.0.tar.gz here] and compile and install it with the included make based installation (see the included README file).
   
  +
== Best human predictions ==
* Via our [https://rostlab.org/owiki/index.php/Debian_repository Debian repository] (package "fastprofkernel").
 
  +
We have predicted all protein pairs in human for which both (C3) or one (C2) protein are sequence-dissimilar to known reliably annotated human interactions. The best 100,000 predictions in either class can be downloaded [http://www.rostlab.org/~hampt/predictions.tar.gz here].
* By manual [ftp://rostlab.org/fastprofkernel/ download] and installation of the binary .deb package (please see the manual of your distribution for how to install .deb packages).
 
 
For installation on non-Debian based systems, you can [ftp://rostlab.org/fastprofkernel/ download] the .tar.gz source package and compile and install it with the included make based installation (see the README file).
 
   
 
== Bugs and Other Issues ==
 
== Bugs and Other Issues ==
Please report bugs and other issues via [https://rostlab.org/bugzilla3/enter_bug.cgi?product=fastprofkernel Bugzilla].
+
Please report bugs and other issues via [https://rostlab.org/bugzilla3/enter_bug.cgi?product=profppikernel Bugzilla].
   
 
== Reference ==
 
== Reference ==

Revision as of 16:10, 19 October 2014

Description

Profppikernel uses an accelerated version of the original profile kernel [1] to automatically train SVM based protein-protein interaction (PPI) prediction models.

The main executable is profppikernel.

It uses profkernel-core (accelarated profile kernel) and trains or applies a SVM based PPI prediction model. It provides two modes of operation. In the first mode, 'training', the user provides evolutionary protein profiles and PPIs between them (labels) as input and profppikernel outputs a folder with all model files required for predictions. In the second mode, 'prediction', the user provides the path to such a model folder and to the query profiles and PPIs. profppikernel then predicts their probability to interact.

Availability

We have included both executables in a .tar.gz source package, together with a sample classification problem. Download it here and compile and install it with the included make based installation (see the included README file).

Best human predictions

We have predicted all protein pairs in human for which both (C3) or one (C2) protein are sequence-dissimilar to known reliably annotated human interactions. The best 100,000 predictions in either class can be downloaded here.

Bugs and Other Issues

Please report bugs and other issues via Bugzilla.

Reference

T. Hamp, T. Goldberg and Rost, B. (2013): Accelerating the Original Profile Kernel. PLoS One, 8(4), e68459

Contact

For questions, please contact hampt@rostlab.org