Classification of scleroderma and normal biopsy data
and identification of possible biomarkers of the
disease
Proceedings of IEEE Symposium on Computational
Intelligence in Bioinformatics and Computational
Biology 2006 (CIBCB2006), page 306--311. Toronto, Ontario, Canada, IEEE Computational Intelligence Society, IEEE Press, (September 2006)DOI: doi:10.1109/CIBCB.2006.330951
Abstract
Scleroderma is an autoimmune disease of the connective
tissues, which thickens and hardens the affected areas.
Recently, researchers have found evidence that genes
are important factors for this disease, and there exist
consistent differences in the patterns of gene
expressions of skin biopsies from affected and
non-affected individuals. In this paper, we apply
genetic programming (GP) on the gene expression data of
scleroderma and normal biopsies to evolve the
classification rules that can differentiate between
them. In these evolved rules, we have found six genes
that have differential gene expression levels in
scleroderma and normal biopsies and thus individually
can classify all the samples correctly. In addition to
these genes, we have also found some simple rules
containing two or more genes that can classify all the
samples perfectly.
Links and resources
Tags
community
@brazovayeye's tags highlighted