To the top

Page Manager: Webmaster
Last update: 9/11/2012 3:13 PM

Tell a friend about this page
Print version

Data convolution and circ… - University of Gothenburg, Sweden Till startsida
To content Read more about how we use cookies on

Data convolution and circadian rhythm impact identification of biomarker genes for ionizing radiation exposure in vivo: concept study on 131I exposure in mouse thyroid

Conference contribution
Authors Britta Langen
Nils Rudqvist
Toshima Z Parris
Johan Spetz
Khalil Helou
Eva Forssell-Aronsson
Published in 15th International Congress of Radiation Research, Kyoto, Japan, May 25-29
Publication year 2015
Published at Institute of Clinical Sciences, Department of Radiation Physics
Institute of Clinical Sciences, Department of Oncology
Sahlgrenska Cancer Center
Language en
Subject categories Radiological physics, Radiation biology, Cell and Molecular Biology, Medical cell biology, Molecular biology


Background: Expression microarrays have been used increasingly for biomarker discovery of genes related to ionizing radiation (IR) exposure, particularly in vivo. However, diurnal variation of gene expression and data convolution from mixed cell populations can hinder biomarker discovery. For one, candidate biomarker genes may underlie circadian rhythmicity and their expression may oscillate affecting their robustness or indicative potential. For the other, significant responses from a specific cell type can be hidden in expression data from mixed cell populations creating bias in results or even precluding biomarker discovery. Aim: To identify biomarkers of IR exposure in thyroid tissue and asses their robustness with regard to circadian rhythm and data convolution. Methods: Female BALB/c nude mice (n=3–4/group) were i.v. injected with 90 kBq 131I, or mock-treated, at 9am, 12pm, or 3pm and killed after 24h. Total RNA was extracted from excised thyroids and subjected to microarray analysis (Illumina platform). Data were processed with Nexus Expression v3.0 (cut-off adjusted P <0.01; log2 ratio ≥0.58). Enriched biological processes (P value <0.05) were categorized after cellular function according to Gene Ontology terms. Data was deconvoluted by cell frequency of follicular cells and C-cells with csSAM using R/Bioconductor. Thyroid mean absorbed dose was calculated as 5.9 Gy using the MIRD formalism. Results: Twenty-five genes responded to 131I in thyroid irrespective of time of day, notably members of the kallikrein (KLK1) gene family, but direction of regulation and fold-change differed distinctly. All KLK1 transcripts were detected in at least one deconvoluted data set, while five additional KLK1 transcripts were detected upon deconvolution. Deconvolution also increased the detection rate of significant transcript regulation and regulated biological processes: DNA integrity, gene expression integrity, and cellular stress were negative in convoluted data, but showed distinct responses in both follicular cells and C-cells. Conclusions: The KLK1 gene family is a promising biomarker candidate that shows robustness of detection. Circadian rhythm and convolution affected the quality and quantity of detected transcriptional responses and we advocate their consideration in the in vivo setting.

Page Manager: Webmaster|Last update: 9/11/2012

The University of Gothenburg uses cookies to provide you with the best possible user experience. By continuing on this website, you approve of our use of cookies.  What are cookies?