Source code The following is a rough guide to how Webminer works, including the perl script that generates the home page ("mkjavascript") and the perl script that provides the search parameters page and generates the final output ("miner"). This material may be distributed as long as no money is charged and this copyright notice is maintained. Webminer and its components are copyright 2000 by Max Heiman.

Webminer consists of the following components:

• mkjavascript
  This perl script reads the names of all experiment files currently available in the Webminer database (i.e., files in the directory "minerdata"). It then groups the experiments into classes and prints out a pre-formatted HTML page ("jsindex.html") with the relevant information integrated. The mkjavascript code is available here.

• jsindex.html
  This HTML file contains the static part of the webminer home page, to be read by "mkjavascript", and includes markers where mkjavascript will fill in the list of currently available datasets. The file is available here.

• miner
  This perl script has two completely different functions. First, when the user clicks the blue arrow on the Webminer home page, this script takes the experiments that were selected and generates a new form asking the user to set cut-off values for each criterion. Second, after the user has completed that form, this script performs the search and returns the list of ORFs matching the given criteria. The script is available here.

• classdescs, exptdescs, and paperrefs
  These three files contain information about experiment classes (such as "Cell Cycle" and "ORF information"), individual experiments (such as "Alpha-factor release, time 0"), and source references for each experiment in the "minerdata" directory. They have the following formats:

  classdescs
  Format: classname (tab) expanded class name (tab) extended description
  Example: cc Cell cycle Time courses of gene expression throughout the cell cycle; transcriptional phenotypes of mutants in regulators of the cell cycle

  exptdescs
  Format: exptname (tab) expanded expt name (tab) extended description (tab) reference code
  Example: cc.alpha_0 alpha-factor release, 0 min Expression profile of cells arrested with alpha-factor spellman1

  paperrefs
  Format: reference code (tab) expanded reference (tab) author's URL (tab) PubMed URL (tab) on-line journal URL
  Example: spellman1 Spellman et al., Mol Biol Cell 1998 http://genome-www.stanfo rd.edu/cellcycle/ http://www4.ncbi.nlm.nih.gov/htbin-post/Entrez/query?db=m&form=6&uid=9843569&dopt=r http://www.molbiolcell.org/cgi/content/full/9/12/3273

• orfs.names, orfs.paths, and orfs.descs
  These three files contain information on all ORF names, pathways, and descriptions made available by the Saccharomyces Genome Database (SGD). They were decatenated from a single file available by going to the SGD home page and following the links "Download Data" -> "tables" -> "ORF descriptions". The full file from SGD is available here.

  The format of each of these files is a plain text file with, on each line, the name of the ORF; a tab; and the corresponding gene name, pathway, or description.

• minerdata
  This directory contains all the experiment files available to Webminer. Each experiment is stored in a separate file, and the file is named in the format "class.experiment" such as "cc.alpha_0" (for the class "cell cycle" and the experiment "Alpha-factor arrest, time 0").

  The format of each of these files is plain text, with each line containing an ORF name; a tab; and the data point for that ORF in the experiment. For example, "YNL279W 28.44". Files in this directory are automatically made available for searching by the user through the Webminer home page.