Skip to content

an active Machine Learning system for Metabolic Network Models development (CMSB 2015 version)

Notifications You must be signed in to change notification settings

robaki/huginnCMSB2015

Repository files navigation

(this is an archival version -- this repo will not be used in further development. Current version of Huginn can be found here: https://github.com/robaki/huginn )

Huginn - an Active Learning system for Metabolic Network Models development

  1. Copyright info: I, Robert Rozanski, the copyright holder of this work (software, figures and data files), release this work into the public domain. This applies worldwide. In some countries this may not be legally possible; if so: I grant anyone the right to use this work for any purpose, without any conditions, unless such conditions are required by law.

  2. Dependencies:

  3. Python 3.4.0 with following modules (most of them should be installed by default): * re * sys * time * copy * random * pickle * traceback * subprocess * multiprocessing

  4. gringo 3.0.5: http://sourceforge.net/projects/potassco/

  5. clasp version 3.0.3: http://sourceforge.net/projects/potassco/

  6. XHAIL (System for eXtended Hybrid Abductive Inductive Learning) https://github.com/stefano-bragaglia/XHAIL Note that ./temp/xhail.sh must be edited to point to gringo and clasp

  7. Usage: To run existing test cases use command ./evaluator.py > log. The process will print many warnings that can be safely ignored. Important information will be saved in the log file as well as in the ./pickled_archives folder. The latter can be read and analysed using functions from the ./development_analysis.py file.

  8. Additional files: ./simulation_data_and_analysis.zip and ./figures.zip contain files used in preparation of a manuscript about Huginn for the 13th conference on Computational Methods for Systems Biology. The paper can be downloaded from here: http://link.springer.com/chapter/10.1007/978-3-319-23401-4_13 (behind a paywall - feel free to contact me to get a copy if your institution doesn't provide you access)

About

an active Machine Learning system for Metabolic Network Models development (CMSB 2015 version)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published