Application of the ROC curve to microarray data (May 26, 2005). Category: Data mining

Life is full of surprises. When I was looking at whether the software package R could compute and analyze Receiver Operating Characteristic (ROC) curves, I found out that there is an application of ROC curves for microarray data. Apparently, the positive false discovery rate can be conceived of in a diagnostic testing format as relating to the positive predictive value. I have not had time to read all the details, but there is a nice paper on this at

• ideas.repec.org/p/bep/mchbio/1030.html

At the same time, I also found a nice paper talking about some of the problems with supervised learning:

• bioinfo.cnio.es/~rdiaz/chapter-azuaje-dopazo.pdf

Oh, and here's another interesting paper about modifications to the Biplot:

• cbis.anu.edu.au/coral/docs/VCGB_125.pdf

When will I ever have time to read all this stuff?

By the way, my search for ROC stuff was sparked by a person who was interested in free software. R will do the trick, but it is a large program that takes a lot of time to learn. Some other choices are documented at:

• www.vetschools.co.uk/EpiVetNet/ROC_analysis_software.htm

This webpage was written by Steve Simon and was last modified on 07/08/2008.