Genome-wide efficient attribute selection for purely epistatic models via Shannon entropy

Manzourolajdad, A, Saraee, MH ORCID:, Mirlohi, A and Javan, A 2008, 'Genome-wide efficient attribute selection for purely epistatic models via Shannon entropy' , International Journal of Business Intelligence and Data Mining, 3 (4) , p. 390.

Full text not available from this repository. (Request a copy)


Epistasis plays an important role in the genetic architecture of common human diseases. Most complex diseases are believed to have multiple contributing loci that often have subtle patterns which make them fairly difficult to find in large data sets. Disorders that follow purely epistatic models cannot be detected by cases/control studies based on individual analysis of susceptible loci. The computational complexity of performing exhaustive searches for detecting such models in genome-wide applications is practically unfeasible. Furthermore, with ever-increasing number of both genotypes and individuals on one side, and little knowledge of complex traits on the other, it is becoming fairly difficult and time consuming to perform systematic genome-wide studies on such traits. We present and discuss a convenient framework for modelling epistasis using information theoretic concepts and algorithms inspired by such an approach. These generalised algorithms, which are especially in favour of purely epistatic models, are applied to both simulated and real data. The real data represents the genotype-phenotype values for Age-Related Macular Degeneration (AMD) disease. Many two-locus purely epistatic patterns were found for AMD. A new visualisation approach is also presented for the purpose of better illustrating epistasy for cases where the number of loci is more than two or three.

Item Type: Article
Themes: Health and Wellbeing
Media, Digital Technology and the Creative Economy
Schools: Schools > School of Computing, Science and Engineering > Salford Innovation Research Centre
Journal or Publication Title: International Journal of Business Intelligence and Data Mining
Publisher: Inderscience
Refereed: Yes
ISSN: 1743-8187
Related URLs:
Depositing User: Prof. Mo Saraee
Date Deposited: 19 Oct 2011 10:38
Last Modified: 27 Aug 2021 20:01

Actions (login required)

Edit record (repository staff only) Edit record (repository staff only)