rmodel source and custom CDFs

Description of rmodel, taken from the manuscript:

Rmodel was developed as a package for the open-source R-project. Rmodel divides a microarray probeset into segments that represent the sequence boundaries of transcribed regions that change by different amounts when comparing two samples, as expected for alternative transcript processing events. Processing events can reflect changes in generation, e.g., alternative polyadenylation (APA) or alternative splicing (AS), or subsequent processing, e.g., degradation or deadenylation, of transcript isoforms. To identify segmentation patterns, rmodel considers all possible subdivisions of a probeset. Subdivisions are evaluated by walking along each probeset in a 5' to 3' direction, evaluating six probes at a time. A modified t-test is calculated from the three probes on either side of the segmentation point. An additional condition was placed on the difference in logratios (rdiff) between the two sides of the segmentation.

Getting and installing rmodel

    Install rmodel version 1.1 from within R [(LINUX versions) R>=2.7, BioConductor>=2.4]
  • source("/rmodel/packages/downloadRmodel.R")
    Install rmodel version 1.0.36 from within R [(all platforms) R>=2.7, BioConductor>=2.4, (windows also requires Rtools installed)]
  • source("/rmodel/packages_windows/downloadRmodel.R")
  • getRmodelAll()
  • or

  • getRmodel()
  • getRmodelMouse()
  • getRmodelHuman()
  • getRmodelZebrafish()

Custom CDFs used with rmodel

Documentation (a work in progress)

Supplemental informaion from the Manuscript

    To re-create these files:
  • install rmodel and the mouse4302rmodelcdf as described above
  • download the oocyte data and R script (hyperlink)
  • place oocyte.tar.gz in new folder
  • unzip the data: tar -zxf oocyte.tar.gz
  • run rmodel from an R window: source("info.R")