
As an Amazon Associate and affiliate partner, Menrva Books earns from qualifying purchases. Learn more
Microarray expression data contain expression levels of a large number of genes and have been used in many scientific research and clinical studies. Due to its high dimensionalities, selecting a small number of genes has shown to be beneficial for many tasks such as building prediction models for a particular disease or gene regulatory network discovery. Traditional gene selection methods, however, fail to take the class distribution into the selection process. In Biomedical science, it is very common to have microarray expression data severely biased having very small number of diseased samples. These biased sample sets require special attention from researchers for identification of genes responsible for a particular disease. In this work, we propose three feature filtering techniques, Higher Weight ReliefF, ReliefF with Differential Minority Repeat and ReliefF with Balanced Minority Repeat to identify genes responsible for fatal diseases from biased microarray expression data. Our solutions will help Bioinformatics, Computer Science and Biomedical Research groups to filter potentially hazardous genes in an efficient way.
This work addresses the challenge of identifying disease-related genes from microarray data that is heavily skewed towards healthy samples. The author introduces three novel feature filtering techniques—Higher Weight ReliefF, ReliefF with Differential Minority Repeat, and ReliefF with Balanced Minority Repeat—designed to overcome the limitations of traditional methods that neglect class distribution. These techniques are specifically developed to handle the common scenario in biomedical science where the number of diseased samples is disproportionately small compared to healthy ones. The proposed solutions aim to provide Bioinformatics, Computer Science, and Biomedical Research groups with an efficient means to filter genes that may be responsible for fatal diseases.
The proposed feature filtering techniques are presented as novel solutions to a specific problem in bioinformatics: handling biased microarray sample sets. The work focuses on the technical aspects of gene selection, aiming to improve the identification of disease-related genes by accounting for imbalanced class distributions. The author's stated goal is to provide practical tools for research groups working with such data, suggesting an intended audience of specialists in the field. The approach emphasizes efficiency and accuracy in filtering potentially hazardous genes.
Page Count:
80
Publication Date:
2009-06-26
Publisher:
VDM Verlag
ISBN-10:
3639170245
ISBN-13:
9783639170245
No comments yet. Be the first to share your thoughts!