Cluster_Boost: Identifying Class Members in Severely Unbalanced Data

Genomic Sweeping for Hypermthylated Genes

Published in Bioinformatics. 2007 Feb 1;23(3):281-8

Motivation: Genes silenced by the aberrant methylation of nearby CpG islands can contribute to the onset or progression of cancer and represent potential biomarkers for diagnosis and prognosis. Relatively few have thus far been validated as hypermethylated in cancer among over 14,000 candidates with promoter region CpG islands. A descriptive set of genes known to be unmethylated in cancer does not exist. This lack of a negative set and a large number of candidates necessitated the development of a new approach to identify novel genes hypermethylated in cancer.

We developed a general method, cluster_boost, that in an imbalanced data setting predicts new minority class members given limited known samples and a large set of unlabeled samples. Synthetic datasets modeled after the hypermethylated genes data show that cluster_boost can successfully identify minority samples within unlabeled data. Using genome sequence features, cluster_boost predicted candidate hypermethylated genes among 14,000 genes of unknown status. In primary ovarian cancers, we determined the methylation status for 15 genes with different levels of support for being hypermethylated. Results indicate cluster_boost can accurately identify novel genes hypermethylated in cancer.


Supplemental Files
Supplemental Table S1 (xls) – List of 63 genes previously reported to be hypermethylated in cancer.

Supplemental Table S2 (xls) – List of 64 sequence features that described promoter regions of known and potentially hypermethylated genes.

Supplemental Table S3 (xls) – Prediction results for 14,249 genes using the cluster_boost algorithm.

Supplemental Figure SF1 (pdf) – Distribution of genes at each classification threshold for 1%, 5%, 10%, and 20% new synthetic samples (SD2)


cluster_boost.tgz – Matlab implementation of the cluster_boost algorithm. See readme.txt for details.


Hypermethylated_vector.txt.gz – Accessions, locations, and feature vectors for 63 known hypermethylated genes.

Unlabeled_vector.txt.gz – Accessions, locations, and feature vectors for 14,249 genes on unknown methylation status.