Publications

Bioinformatics world, a quarterly publication, discussed data mining applications in bioinformatics.

Bioinformatics is at the interface between biology, medicine, mathematics, and computer science. The major challenge for this research is to elucidate the relationship between sequence, structure, and function ... An example of data mining of DNA, is the "peculiarity oriented mining" proposed by researchers at Maebashi Institute of Technology and Waseda University in Japan. This algorithm is based on the proposition that if a sequence is peculiar then it might also be interesting. They define data to be "peculiar" if it is very different from other objects in the data set. A peculiarity factor is assigned to each object by calculating the sum of the square root of the conceptual distance between that object and other objects in the set. The peculiar data is then studied by assigning a "relevance" to each piece of data. This technique has revealed "reasonable and interesting", albeit specialised and abstruse, relationships between amino-acid models and experimental results.

See full article at http://www.bioinformaticsworld.info/feature3a.html