Machine Learning of Functional Class from Phenotypic Data

Used in the papers:

  • Clare, A. and King R.D. (2001) Knowledge Discovery in Multi-Label Phenotype Data. In proceedings of ECML/PKDD 2001. abstract
  • Clare, A. and King R.D. (2002) Machine learning of functional class from phenotype data. Bioinformatics 18(1) 160-166. abstract
(See papers, code for C4.5 modifications).

Data

Original sources of data: Reformatted for C4.5 (large files are bzip2-ed): Description of attributes (growth media).

Functional categorisation (hierarchy) as bzipped prolog-format file: here taken from MIPS on 27 September 1999. Current (possibly more up-to-date) categorisation available from MIPS here.

Tab separated class list file for conversion between class names and class numbering scheme. For example, class 9/1/0/0 is "biogenesis of cell wall (cell envelope)". This is the first subclass of the more general class 9/0/0/0 "CELLULAR BIOGENESIS".

Annotated rulesets

Prediction for orfs of unknown function