12.07.2015 Views

View - ResearchGate

View - ResearchGate

View - ResearchGate

SHOW MORE
SHOW LESS
  • No tags were found...

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Modeling Transcription Factor Target Promoters 147The user should run the Random Forest program repeatedly, each time droppinga certain percentage (e.g., 20%) of the least important variables. Finally, the usershould focus on the top ranking variables for constructing the CART tree. Evenif some of these top ranking variables are not included in the initial CART tree,running CART repeatedly by selecting different combinations of the variableswould give a better idea of possible TFBS modules. The list of ranked featurevariables (Fig. 2) is valuable information for prioritizing the TFs for furtherexperimental verification.4. The following sequence of R commands ranks the variables in the decreasingorder of mean decrease in accuracy (rf.imp[,3]) and Gini index (rf.imp[,4]), andselects the union of top 10 ranking variables either according to mean decrease inaccuracy or Gini index.a. rank1

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!