Illinois Genetic Algorithms Laboratory

ComputerWorld has recently published an article on using Microsoft’s anti-spam software in developing a vaccine to fight HIV:

Spam-filtering technology may soon save millions of lives, thanks to the technology’s potential use in developing a vaccine to fight the deadly human immunodeficiency virus (HIV).

Researchers are conducting in vitro tests of HIV vaccine models developed using Microsoft’s anti-spam software, according to Kevin Schofield, general manager, Microsoft Research in Redmond, Washington.

The project, he said, is a joint initiative between Microsoft Research, the University of Washington (UW) in Seattle, and the Royal Perth Hospital in Australia.

The great similarity between how spam works and how HIV cells mutate in the human body has allowed researchers to use Microsoft’s machine learning and data mining algorithms to analyze the virus’s genetic sequences. The purpose was to identify patterns within the genetic mutations of the virus and the patient’s immune system, according to Schofield.

Several years ago, a different approach in the same area was pursued by Stefan Kramer, Luc De Raedt, and Christoph Helma, who analyzed the AIDS Antiviral Screening Database using the levelwise version space algorithm that forms the basis of the inductive query and database system MOLFEA (Molecular Feature Miner); for more information about this approach, click here. Stefan Kramer later coauthored a paper by Ulrich Rueckert, where they present another approach to the same problem, which is based on mining for free (unrooted) trees (more info can be found here).