Email updates

Keep up to date with the latest news and content from BMC Genomics and BioMed Central.

Open Access Research article

Time-series clustering of gene expression in irradiated and bystander fibroblasts: an application of FBPA clustering

Shanaz A Ghandhi1, Anshu Sinha2, Marianthi Markatou34 and Sally A Amundson1*

Author Affiliations

1 Center for Radiological Research, Columbia University, VC11-215, 630 West 168th Street, New York, NY, 10032, USA

2 Department of Biomedical Informatics, Columbia University, New York, NY 10032, USA

3 Department of Biostatistics, Columbia University, New York, NY 10032, USA

4 Department of Statistical Sciences, Cornell University, 301 Malott Hall, Ithaca, NY 14853-3801, USA

For all author emails, please log on.

BMC Genomics 2011, 12:2  doi:10.1186/1471-2164-12-2

Published: 4 January 2011

Abstract

Background

The radiation bystander effect is an important component of the overall biological response of tissues and organisms to ionizing radiation, but the signaling mechanisms between irradiated and non-irradiated bystander cells are not fully understood. In this study, we measured a time-series of gene expression after α-particle irradiation and applied the Feature Based Partitioning around medoids Algorithm (FBPA), a new clustering method suitable for sparse time series, to identify signaling modules that act in concert in the response to direct irradiation and bystander signaling. We compared our results with those of an alternate clustering method, Short Time series Expression Miner (STEM).

Results

While computational evaluations of both clustering results were similar, FBPA provided more biological insight. After irradiation, gene clusters were enriched for signal transduction, cell cycle/cell death and inflammation/immunity processes; but only FBPA separated clusters by function. In bystanders, gene clusters were enriched for cell communication/motility, signal transduction and inflammation processes; but biological functions did not separate as clearly with either clustering method as they did in irradiated samples. Network analysis confirmed p53 and NF-κB transcription factor-regulated gene clusters in irradiated and bystander cells and suggested novel regulators, such as KDM5B/JARID1B (lysine (K)-specific demethylase 5B) and HDACs (histone deacetylases), which could epigenetically coordinate gene expression after irradiation.

Conclusions

In this study, we have shown that a new time series clustering method, FBPA, can provide new leads to the mechanisms regulating the dynamic cellular response to radiation. The findings implicate epigenetic control of gene expression in addition to transcription factor networks.